a system composed of people can be debugged in the same manner as code

False.

o It's point to point; I start where I am and I end up where I'm going

It's nice to have valet parking wherever you go, but I usually have to park and walk somewhere and then walk to the entrance.

o It's considerably more secure; windows up, doors locked, only trusted riders are on-board, and I control the vehicle

If only that were enough to make cars safer than buses and trains.

o I have my music (and my ham radio gear), in short, the environment is customized for me

With all those distractions, you're asking for trouble!

o There's no waiting, no calling, and no communications problems

There's also no reading or texting and you can't use your laptop while you commute. It's all dead time.

I think the most common case by far is that people use it because they have to use it.

People must not like driving very much either or cities wouldn't need to force property owners to have parking (why do we manage parking the same way the Soviet Union managed toilet paper and other basic provisions?), and the roads would pay for themselves 100% through gas taxes and user fees instead of less than 50%.

When you say the 85th percentile rule "has been properly and professionally researched and reproduced," I believe you are referring to the Solomon Curve published in 1964. But that was back when people drove more safely out of necessity because cars were deathtraps. Do you have any evidence that the 85th percentile rule reduces fatalities today? Bonus points if you can prove that it reduces car-motorcycle, car-bicycle, and car-pedestrian fatalities. Good luck!

It's unfortunate when speeds are set by the fastest 15% of drivers in a country where 80% of drivers think they are above average and are therefore unqualified to determine a reasonable and prudent speed.

speed limits are almost *always* way lower than the speed it's actually safe to drive.

That's partially true, assuming a very broad interpretation of the word "almost". Speed limits are set according to ideal driving conditions (daytime, dry roads, perfect visibility, drivers paying attention to the road) with a small margin for safety. Under anything other than those ideal conditions, the Basic Speed Law requires (see link above) you to slow down to a reasonable and appropriate speed despite the posted speed limit.

The flashing red is the yellow phase.

Where is that true? Not in California.

It's probably the same thinking that gave pedestrians two red phases (flashing and solid "don't walk"), one green phase, and no yellow phase.

But building more roads doesn't make people drive more...

Actually, yes, it does.

People aren't going to "fill up" the freeway at times when they don't have a reason to be on it.

Your logic is circular. That's funny.

Anyway, whether a freeway fills up depends on the capacity of the freeway and the number of people who want to take their cars on it at the same time. You don't seem to believe it, but this is always true, even in the middle of the night.

You aren't going to pay people vast sums to incentivise driving in the middle of the night to keep your traffic flow constant.

Why would you need to? Just wait long enough without widening a freeway and eventually it will fill up 24/7. If you use express tolling to prevent traffic congestion, it will never get congested and therefore it will never need to be widened to eliminate traffic congestion, even though it's moving 100% of its daily capacity.

Please elaborate.

If any highway anywhere moves half of its peak capacity in a day, the transportation engineer responsible for it should be fired.

But doesn't a demand curve show a way to keep traffic demand, and therefore traffic flow, constant, even 24/7?

And because the fiscally optimal amount of road is the amount where the marginal cost of adding a lane equals the marginal revenue from adding it (MC=MR), doesn't this prove that a nonzero amount of traffic congestion on an unpriced road is optimal?

The idea that you may not need more roads is... completely foreign.

Perhaps the hamburger analogy will help:

Let's give everyone free McDonald's hamburgers. Let's put 10,000 hamburgers a day on a table in front of the Capitol (or wherever).

What would happen? People would take and eat the hamburgers, and once word got out, all 10,000 hamburgers would be taken very quickly every day. We may thus infer that because people need food and they really seemed to like those burgers, McDonald's hamburgers are an important public good.

A city planner might notice a problem: those 10,000 hamburgers just aren't enough. They get taken very early in the morning, so not everybody has a chance to get a hamburger. The obvious solution -- because burgers are a highly-valued public good -- is to provide more free burgers. So the city planner starts to provide 20,000 hamburgers a day.

You can see where this is going. People start going out of their way to get the free hamburgers, and planning their day around that trip. The city has to keep providing more and more free burgers -- eventually millions a day -- to keep satisfying the demand for free hamburgers. The competing food markets crater, because who would pay $2/lb for apples when you can get as many free burgers as you want (although maybe you have to wait in a 30-minute line). Public health goes to hell, because everybody's eating six burgers a day. And yet, everybody likes their free burgers and the Hamburger Department is an untouchable political powerhouse. Proposals for a 10-cent hamburger fee to cover the huge costs of hamburger provision get shot down by public outrage.

What's the problem here? The problem is that food is indeed a necessity, and yes, people seem to like McDonald's hamburgers -- but the fact that people will take free burgers does not prove that they are "highly valued" by the market. We are not seeing actual demand for burgers. We are seeing induced demand for a good which is being provided at artificially low prices.

But for some reason, replace hamburgers with roads and everybody goes nuts.

In short, the fact that a new lane or road immediately fills up with traffic does not "prove" that there was a high demand for that road -- it proves that people will use way too much of something that's free.

Personally outside of some of the most rural parts of the US I've NEVER seen "empty highways".

Mostly empty highways then. Can you name one that moves even half of its daily capacity (peak hour capacity times 24)? Good luck! Ha ha!

They're putting an expensive new tolling system on 405's commuter lane that will dynamically increase tolls in response to increases traffic so that it stays clear for busses... Of course, that's actually going to make the normal 405 traffic *worse*, because they're simply pushing the traffic into the normal lanes.

That's most definitely false, and here is proof:

On the S.R. 91 Express Lanes, vehicle throughput during the most congested hours exceeds 1,600 vehicles per lane per hour on the priced lanes, while the adjacent free lanes carry only about 800 vehicles per lane per hour.

So as you can see, the express lanes move twice the auto traffic as the regular lanes, not less.

No, Washington doesn't need more lanes on the 405 at this time. You only need to manage the existing lanes better.