I don't think either of the following have been established:

1. We should expect existing lax enforcement to have a significant effect on speed.

2. Existing lax enforcement has a nonexistent effect compared to the baseline (no enforcement at all).

Even if you established both of these, it still wouldn't be sufficient to make the logical inference that rigorous enforcement would be ineffective for decreasing speeds. It would not be at all surprising that behavior goes unchanged if enforcement rates are below 0.1%. It would be very surprising if enforcement rates closer to 10% had no power to change behavior.

The problem is that it's really an economic problem. The existing enforcement is a revenue generation method -- everybody exceeds the speed limit, so you can put a cop anywhere who can just sit there and hand out tickets. It's net income for the government, but there isn't enough enforcement to actually change behavior.

If you added more police then people would stop speeding, but then people stop speeding -- there is no more ticket revenue. Now how are you paying for all the traffic cops?

That's an interesting technical problem. PID controllers answer a similar problem in control theory. I don't have the answer here, and I'm not certain there is one. It really depends on how many more cops we'd need to get to a state where enforcement can be rigorous, and whether we as a society prefer the tradeoff of paying for those enforcers to get safer roads. Tough question. But I like tricky questions aimed in a true direction better than easy, wrong answers like, "ticketing people has no effect on speeding."

If people were to stop speeding, perhaps there'd emerge a culture in which speeding is considered bad, reducing the need for traffic cops.

As soon as you stopped enforcing, people would speed again.

In cultures where e.g. drunk driving is considered an awful thing to do, people mostly self-regulate - there's some policing necessary, but it's not like the police are trying to stop something everyone wants to do. There's no reason it would be impossible to create a culture in which speeding is treated similarly.

There's quite a few countries in which people avoid speeding, even when there's no cameras around.

Aggressive enforcement isn't how you do that. And it's a lot harder to do it with speeding than drunk driving because it actually is a lot more dangerous to drive drunk than to drive the 85th percentile traffic speed even though the speed limit is set 10MPH lower than that.

In most of the existing places with a culture of respect for the speed limit, they have a higher speed limit.

Fair enough. The article may not have sufficiently established such a thing. But enforcement when it works is likely to only work when enforcement is frequent enough to create an expectation of getting caught. Given that this would require a significant increase in manpower to truly be effective I'm not sure it's sustainable.

I suppose you could find a way to automate it similar to red light cameras but even those have been shown to be rather ineffective at actually lowering the incidence of violations so I'm not sure anything less than the presence of a police vehicle with flashing lights that pulls up behind you would have an effect.