Thanks, I agree with most of your comments, especially the last paragraph acknowledging the trade-off between costs and benefits of an inspection policy. The choice of inspections every two years itself represents a decision to give up potential benefits of more frequent inspections.

people complain but generally agree that it is nice to know that your car has everything fixed and every other car is in good condition.

This sounds like wishful thinking. In order to have "everything fixed" we have to assume that every possible safety defect is included in the inspection standards, that every inspector does a perfect job, and that no one evades or cheats the system. Even so, two years is plenty of time for things to go wrong between inspections.

Obviously the costs would be too high to aim for perfection, so we live with a balance of costs and benefits of imperfect inspections. One possible unintended cost might be encouraging drivers to assume their vehicles are safe as long as they pass inspections, so they are less inclined to look for problems on their own.

The fact that we can find arguments for both sides online doesn't mean that both arguments are equally strong. If inspections significantly improve safety, it should be possible to measure that improvement in states with inspections. If we can't find clear evidence of a net benefit, we should not claim that there is a net benefit.

I also wonder why I no longer see broken down cars at the side of the road in the U.S. It used to be common to stop and see if stranded motorists needed a hand. I think cars are generally more reliable, drivers are less inclined to try and fix problems, and most people can quickly summon help with a mobile phone. Flat tires used to be common. In the 1970s, "typical bias ply tires lasted less than 20,000 miles and were only expected to be in service for about two years" but today's tires are "approaching 80,000 miles of treadwear."