ORIGINAL: ladner
One thing I have noticed in many of these games, and back when I used to play War in Russia, I was guilty of having factories produce all Fw-190s instead of Me-109s, since the Fw-190 was a better plane, at least in game statistic terms. In a lot of these games this type of thing occurs, for example if you are Britain, why bother with the Hurricane, when the Spitfire is so much better.
There is already a solution to this in World in Flames. It's only part of it but the idea there is that you get a mix of units that you choose from randomly. You build a fighter and if lucky in 1941 you got a FW190 and if not you can still get an old Me-109. As there are more of the later at the beginning the chances are not even even.
I believe it would be easy to give a table by type of units with a percentage by subtype. Like in 1939 you have 75% change of getting a Me-109e, 15% of a less version and 10 percent of an old plane that I forgot the name but was still used. For the bomber you would do 60% Me111, 15 Ju-88, 20 Do-17 and 5% Condor or something like it. See the idea.
Then with your choice in research those %age will vary. Research Jet Engine since the beginning and you can get 5% Me262 in 1941 then 15% in 1942 and so while you scrap the older model.
You could even choose to scrap voluntarily older model at a cost for as you say restructuring the industry.
That way you don't end up with always the best planes but still a realistic variety with enough influence at a cost.
Now on my sense at a strategic level the subtype of plane is not that important but more the numbers and the type but of course it gives some flavors to the game.