That's something you'll need to backup with facts. In the past 10 years where I've lived various governments have caved under pressure to let kids get their L plates at an earlier and earlier age.
With the reference to L plates, perhaps you're British? Here is a Guardian article with some statistics in the first paragraph about decline in licensure among 17-to-20 year olds, as well as 21-29. Here's a similar set of statistics for the US.
That would have a lot to do with very little information being passed onto them and people complaining about it everytime they do. How they would love to know how many hours you spend behind the wheel, as you already alluded to :-)
So we're in agreement - without that data they can't do much more than average across the population. But that unfairly (for some definition of "fair") benefits some people while punishing others, assuming you believe in some notion of the intrinsic safety of a driver
Not only did it count for me, the insurance company promoted the classes and I was able to claim back the cost of the class from the insurance company.
This wasn't a class the public could take - it was a class about emergency driving, with lights and siren. It did involve going on a skid pad and learning how to drive through a loss of traction, as well as slalom and reverse-slalom as well as general situational awareness (there's no rear window so you have to track where nearby cars are). Most useful to me was learning driver "psychology" as it were, learning how people in aggregate respond to unusual situations and seeing lots of examples of the ways drivers can screw up given a surprising event means I'm rarely surprised by what someone on the road does. I've had to take evasive action several times to avoid an imminent crash and it's certainly helped to know the limits of the vehicle performance, the road surface, and what the other driver(s) are likely to do given the circumstances.
I don't expect the insurance company to promote or pay for such a class, and in fact they would have no business doing so, but if they took it into consideration it would be a sign that they were willing to individualize their notion of driver risk. But they aren't interested.
Really to be fair, flying a plane is a very different skill set than driving a car. It is a much more refined skill with a metric shitload of inference based on information provided by instrumentation. Where looking out the window becomes important a lot of information is incredibly subtle (at height the landscape can appear almost unmoving) By comparison one of the biggest problems with new drivers is they spend too much time looking at instruments in a scenario where pretty much anything can jump out infront of their windscreen at any moment. It's a very different kind of situational awareness, and personally I don't believe that being a pilot would make you a better (or worse) driver on the road but I would be happy to see some stats to correct me.
The biggest problem with new pilots is that they spend too much time looking at instruments, too. Most private flying is done visually and "seat of the pants", and a flight instructor will commonly cover up all the instruments if a new student is fixating on something (usually the artificial horizon) to try to fly the plane without a "feel" for it. We don't typically fly high enough for the landscape to seem still; it's typical for me to fly at 3500' or 5500' and I spent a lot of time lower than 2500'.
I never said that they were exactly the same skillset, and I don't have any data, but becoming a pilot forces you to become very very good at multitasking, risk management, planning ahead (both before you get in the plane and figuring out what you can do in spare time to keep ahead of the situation), and monitoring the environment.
Let me give you an example: if the weather is good enough (which it is, if you're flying visually) it's on each pilot to "see and avoid" other airplanes. But in controlled airspace, it's a good idea (and sometimes required) to be monitoring the appropriate radio frequency where air traffic controllers are communicating with other airplanes sharing the same few-thousand square mile sector as you. Similarly at an untowered airport there's an advisory frequency that pilots are supposed to self-announce their position and intentions. So you get very good at building a mental model of the airspace around you and where everybody is in 3D space and where they're likely to go, since building this model makes it easier to spot them and make sure you don't fly into each other. Even at a towered airport, where the control tower is helping you find everyone you're flying around with (e.g. if you're practicing landings there might be 5 or more planes "in the pattern"), you have to keep track of where everyone is and make sure there's enough space.
How does this relate to driving? Well, this is about a hundred times harder than keeping track of the cars in front, behind, and in the adjoining lanes and making sure you don't change lanes into a car in your blind spot. And it's very good practice for that subconscious "map of the world" stuff. There are serious differences, in particular that during flying you overwhelmingly can get away with a few seconds of not looking out the window or touching the controls, and in fact it's expected, whereas such a mistake would be frequently costly while driving. But I can't think of a way that learning to fly would make you less safe of a driver (except perhaps by cockiness, which is pretty frowned upon in the aviation community), and considering that there are several areas of overlap that learning to fly improves your skills, I'd say it would make you safer. I certainly feel a lot more "on top of" the situation while I'm driving, and see close calls coming from further away, and am better-prepared to respond sensibly.
Totally agreed on the driver training aspect. Frankly people on the road scare me, and I've seen the aftermath of hundreds of crashes. I see no particular reason to trust the average US driver, and defensive driving only gets you so far. Nine years and two states ago I spent 10 minutes showing a guy (who spent the entire time filling out paperwork) that I could drive, and nobody's questioned me since - or will again for a very long time. But it's totally legal to have a license for 20 years and have driven about 10 hours total, and none in the past 5 years, and then to rent a car and go on the highway opposite direction to you at 70 miles an hour. If that's not an argument for raising the bar for initial licensure, I don't know what is.