It's not about hardware anymore, if it ever was. It's about the entire platform and the software is the major part of it. Not at all different to how they currently operate, essentially locked into the windows platform, and they usually see it as a benefit, not a drawback. That's why it's so hard for multivendor approaches, like linux-based systems to catch on in the corporate world.

1. Why would the phone power that thing? If anything, it of course includes its own charger and battery and changes the phone, while it's docked.
2. That's what we have now, and requires maintenance of two separate systems: the phone and the laptop. Unifying them would definitely be a benefit, not a drawback.

Why would a single hardware provider be worse than a single software provider? The latter was never an issue for most companies. If anything, it's better for them if there is just party to support for both hardware and software if something goes wrong. You know, most companies aren't hackerspaces, where every user spends all their time tinkering various devices just for the sake of tinkering.

You can likewise get "any" bandwidth from any other connector too, but without the lossiness of the analog signal. The "VGA" signal itself is just really analog RGB with h/vsync and optional EDID information. Nothing specific about that either. There are also hard limits on what you can push through a setup like that. Definitely less information in most cases than on the more modern digital video buses. A spec defines only the minimum a specification-compatible device should support, just like in the case of VGA, every VGA card should be able to output at least 640x480@60Hz, just like a single-link DVI should be able to drive at least 1920x1200@60Hz, but there are really no limits on that design either, it's just up to cabling there too before the signal-to-noise ratio becomes too bad. HDMI itself is practically the same as DVI, but with added encryption and audio signals. Newer-spec HDMI ports can be still used as "over-spec" clocked single-link DVI ports with just passive adapters that wire one type of connector to another, which is useful for monitors like the IBM T221, which supports much higher bandwidths than regular single-link DVI can provide, but older versions of it didn't support dual-link DVI either, because it wasn't specified yet back then.

D-Sub itself doesn't mean anything specific. It's just a short for "D-subminiature" type of of connector. In the case of VGA is the horribly designed connector with 15 pins crammed in the footprint where a typical D-connector would have 9 pins (like on rs232 com-ports on legacy PC's and atari joystick ports). If you want to call the connector something, at least use the full name, D-sub-15, because there are all kinds of pin counts for D-sub connectors as well as regular D-connectors. A dense setup worsens the signal quality somewhat due to cross-talk (results in ghosting on the screen) and makes the connector itself weaker than in would otherwise be. It's not rare to get an occasional bent pin on the connector, because it doesn't really take any effort. Regular D15 connectors were used on at least the old-school Macs for their RGB output. Other systems used other connectors, ranging from some really exotic ones to just plain multiple BNC connectors. On legacy PC's the same regular D15 connector type was common for the (analog) joystick port.

If resolutions go up, the bpp could actually be brought down. Dithering won't be visually distinguishable even at let's say 4bpp @ 600ppi. At 100ppi it's definitely distinguishable, of course, because you can distinguish each pixel.

Laptop displays aren't more expensive than desktop displays. They are the same kind of panels, just in a different package.

They can still inflate screen sizes even if they go with higher resolutions. The limit will still be basically what people are willing to pay. It'll take a while before wall-sized 1200ppi displays are a commodity.

I do. Photographic images are antialiased by default. Try switching off line-drawing images and suddenly you'll require at least 400-600ppi at regular viewing distances to make the pixels "barely visible". Same for text, which is technically a bunch of line-drawn glyphs. I'm using a 200ppi monitor here and I still need anti-aliasing to make pixels "barely visible".

Not any resolution at any refresh rate. VGA is limited by the display adapter's DAC's speed, which varies from one display adapter to another. Likewise, there are no limits on resolution on DVI, HDMI or DisplayPort either, the limit is in the bandwidth, and there are variation there too. Also, to nitpick, VGA is strictly speaking just 640x480@60Hz. What you are speaking of are nonstandard, hot-rodded VGA signals, something the spec was never intended to drive and the signal is transferred lossy, whereas the digital interfaces are lossless.

It's not bad for video. You can go down to 30Hz, 24Hz or 25Hz to match the video frame rate exactly, if you must. Nothing prevents you from attaching lower res, higher refresh displays and TV's for some games and movies, however. I love playing games in high-res, even if they are limited by the display refresh rate. The monitors are also overclockable up to 60Hz refresh rates (I haven't tried it myself yet). The overclockable parts are the huge FPGA chips driving the panel circuits. To give enough bandwidth to drive them at those refresh rates, you need a pair of HDMI 1.3 outputs with HDMI-DVI adapters or a DisplayPort output with a dual DVI adapter.

You don't know what you are talking about. TFT displays don't flicker even at 1Hz. The contrast is great, so is the colours (8bpp), it has more brightness than I need, I commonly use the range between minimum and half brightness. Low refresh rates on TFT's just translates to a bit more perceived lag for content, but 43Hz isn't bad and I run mine off two DVI outputs resulting in 34Hz, which is good enough even for movies, which are 24Hz to 30Hz anyway.

I tried it when I installed OS X 10.7 Lion, but went back to the regular mode. I don't want big-ass widgets, I want more workspace. All apps don't support the HiDPI modes, but it doesn't matter, because what they draw is just simply scaled up. Here are two screenshots I took from my setup to illustrate what it looks like and how ready the support was at 10.7.0: http://sorsacode.com/3840x2400_HiDPI.png
Also nicely illustrates the difference between 1920x1200, which is what Safari thinks the resolution is, and 3840x2400 HiDPI, which the other apps on screen supports. However, I just like smaller widgets better, which looks like this: http://sorsacode.com/3840x2400_regular.png
The screen shots are taken on a 22" 3840x2400 VP2290b (rebadged IBM T221) display driven by two single-link heads, hence the menubar split in the middle.

Correct. Typical speckles of dust are a bigger than pixels at 200ppi. Not really an issue and I bet most complainers have large enough dust particles on even their low-res displays covering several pixels at the moment.

Those lines weren't really much thicker than the lines between each sub-pixel on current low-res TFT's. It would take a long time to find them, if you weren't looking, and when you found them, they'd be just as obvious as a stuck or dead pixel on a TFT, even if you hadn't noticed it before.

Sorry, 495MB/sec for the latter.