First thing - most publishers / developers cannot accept ideas submitted to them, and have a policy in place to destroy any idea submissions without reading their details. Reason behind this is they don't want to get sued later for making a game that seems similar to one someone submitted without credit - even if the company had been making the game for a few years at that point. So making a game concept or demo and sending it to your favorite publisher isn't the way to go.

The best way to get your game noticed is to make your game (or at least a demo of it) the way you want, and post about it online. Publishers and developers are always looking for talented folks. If you focus on sound design, great! Show it off! Can you do some cool trick with dynamic clouds and shadows in your level? Or have a really neat style of gameplay? All these things can get the attention of the folks with the money. Valve and other publishers have been known to either take an indie developer or team and either hire them to make an upgraded version of the title, or help them publish, just because they like the game.

Yes, there is a chance that someone will steal your idea and make their own game off of it. If you're super worried about that, then wait until your game is nearly complete before you post anything about it. Fact is, if the game or some idea in it is great, it will be copied. Happens all the time in professional games, too. Bullet time, anyone?

Another great venue to get noticed is game competitions, such as the IGF. Submit your game to these, and you can bet that publishers will see you, your ideas, and think about if your game might fit in their lineup.

As for tools and engines, they're secondary, in a sense. There are plenty of free engines, some dedicated to certain gameplay styles (UDK), others more freeform (Ogre). If you're making a full game that you want to sell, make sure that's not prohibited in the terms of the engine you're working in. Otherwise, treating it as a proof-of-concept demo is the way I'd go, and for that, whatever engine lets you do what you want is the ideal engine. It helps to have your game not look terrible, but there are tons of freelance artists, sound designers, and so on out there to help you with that. Unless, of course, you're trying to show off some new procedural animation gameplay system or something, then make sure you have some nice looking models to back it up!

"we're sitting there late, then I've screwed up." To me, that's the difference between a good manager (one who schedules well, but is sometimes wrong) and a bad manager (one who sees overtime as a "necessary evil"). Yes, sometimes things go badly, and the team or members thereof have to stay late to get things done. It's a pain, it's bad, but it happens. But I've had some managers that make schedules assuming overtime as a part of getting the project done. That, to me, is wrong. Especially since, as you say, sometimes someone has to be there to say when everyone's brain is fried and they need to go home. Usually, it's the manager making this call, and if they've already decided that overtime is good for project sceduling, well...

Well, right now we definitely have no "cure" for cancer - we have ways of killing it, but nothing to make those cells that are cancerous to go back to being normal, productive, healthy, happy cells. I very much doubt we'll ever find a way to "cure" cancer, once it has started in a body. Prevent, yes, treat/kill, yes. Cure, no.

That said, I'm all for new and better ways of killing cancer cells, preferably while leaving the patient healthier during the process. And if we can prevent it before it starts, even better!

Professional game programmer here, I've worked at (and hired for) several large companies, so I've got a fair idea on what they look for, at least on the programming end.

First, as several people have pointed out, design and programming are different aspects of game development. Really, it's broken down into a ton of areas that you can aim for, which include (but are definitely not limited to!):
* Design / Scripting - these are the people who use an engine to make boss fights happen, in-game cinematics, things of that nature. Mostly the player interaction work.
* Design / Level - these are the people who use an engine to lay out the level you play in. They set up doors, lay out camera tracks, etc. Often combined with Design / Scripting
* Design / World - this is more of a writing position. You come up with the world, the people in it, etc., and others fill in the blanks. Usually the lead does this, or lead + combination of others.

* Programming - lots of fields here. UI, tools, engine, AI, sound, networking, graphics, the list is rather long. Out of college, you'll likely start out as a gameplay or generalist coder, unless you've specialized or someone has needs for you to specialize.

* Art / 3D modeler - build models in 3D Studio Max or Maya. Expensive, but more or less the standard at most companies. Often this role is just working on the geometry, but sometimes combined with other roles.
* Art / textures - apply textures to the above geometry. You have to figure out how to lay out the texture for optimal use, know where to use alpha/translucency, mirroring, and other effects.
* Art / skinning & animation - take that geometry, then attach the vertices to bones so the character can walk around.
* Art / concept - draw out sketches and paintings to show what the game and characters can be.

* Q/A / Test - testing a game is not that similar to playing a game. You're going through the same broken area over and over again to figure out exactly what is causing the problem. You try to write it down, or take a video or screenshots, then you have to do it all over again with the next build to see if the problem is really fixed, or mutated into something else. Really good testers are hard to come by, because most people don't want to stay in Q/A. But those that stick it out are extremely valuable.
* Q/A / Lead - come up with plans on how the game should be tested. You need to work with the designers and programmers to figure out when parts of the game will start working, plan out your resources (Testers) so that they get exposure to different areas of the game so that individuals don't get locked into their way of playing, make sure that as much of the game can be tested in a limited timeframe, etc. Again, great Q/A Leads are super-rare.

* Production / Assistant - help out by keeping track of deadlines, assets, making sure that people are on target for their tasks. When something goes wrong, help figure out how this impacts the schedule. Take screenshots, videos, and so on for marketing.

Check out the credits on some of your favorite games. You might find areas of development that you'd like to try out.

If you want to get into the programming side, I would recommend getting a bachelor's degree in Computer Science. Not one of the game development degrees that some places offer, but a straight-up CS degree. That gives you a solid foundation to build your game skills on - which you should do in your spare time. There are a ton of websites dedicated to game development, and the things discussed therein will make more and more sense as you progress through your degree path.

When you apply for a job, be sure to bring home projects to the interview, or submit them with your resume. These show what you can do, but also that you have passion for the job. If possible, have a variety to show off - tests to try out interesting ideas, simple games that are complete (main menu, levels, player victory, etc.), that sort of thing. If you're aiming for, say, a graphics position, be sure to bring in a sample or two that do interesting graphical effects. If you're aiming for a design role, build a level in an established game editor, like Quake, NeverWinter Nights, or Left 4 Dead.

And finally, finish your degree! This shows that you have the guts to stick out a long project, which you'll need to stick through a game project from start to finish. It'll also help if you someday decide to go into a more traditional job, which some game developers do. Game development is a lot of fun, but it's also a lot of hard work, and sometimes a lot of long hours. Hard to get to kids' baseball games if you have a deadline to hit.

That is a pretty major issue you happen to bring up - benefits.
Consider party 1 - a man and woman with a little certificate saying that they're going to stay with each other for as long as they both may live, or at least until the divorce settlement.
Consider party 2 - two people of the same sex, minus the certificate, but wanting the same sort of relationship.

If one of party 1 gets hired by a company like Google, the other member can get health insurance and other benefits. But, this only works if you have that little certificate. Party 2 can't get that certificate. Therefore, the worker's partner can't get benefits. Therefore, they need a higher salary to pay for those benefits.

So, you're already automatically locking out some of your workforce by salary requirements.

Next, imagine you have two choices for a place to work, in different countries, and you and your partner like to read. In one of those countries, they've started making laws about what people who like to read can and can't do. In the other, there are no such laws, or if there are, they embrace people who like to read.

Where do you want to go work?

I'll bite, though I doubt this will be read by you, or honestly anyone.

How long until we go back to the drawing board? Who's to say we ever left it? That's the thing of scientific models and the whole thing about a hypothesis - you keep observing, making adjustments to your ideas, testing them some more, wash, rinse, repeat.

The major problem with proving/disproving climate change/global warming is that a) we're inside the running event of the planet, so it's hard to make outside observations, and b) we can't experiment. So yes, we today say "here's what we've been seeing. Here's why we think it's happening, and this is what that would mean for the future". Tomorrow, we'll say "gee, and this is what we actually saw. Let's adjust our thinking, and see what predicitons we can make from that".

Please keep in mind that current scientific theories are not what you hear spouted on TV, and even those are more advanced than what most people have in mind when they talk about climate change and global warming.