Even though the idea (theory) is very interesting and looks promising because it is focused to only one aspect, I still think it is not practical and will be extremely difficult (and costly) to operate. Why? It is much easier to check up/force the cause (selling vehicles) than to patch the end (vehicle consumers). There is much smaller number to deal with. If the responsibility is pushed down to the consumers, what else do you think would happen? Who is keeping the records? How to deal with them especially when there is an issue? How many people would be involved in doing so? How much it cost to implement, install, train, etc., the system? Many things will happen in both advantages and disadvantages.
Could individuals cheat by flashing their EFI control modules etc before they have their test and then putting it back after, well yes but they can do that kinda stuff in places that do checks now anyway. The smallish number of individuals that cheat won't amount to much.
Again, in theory, you are thinking of the whole situation in a controlled environment. However, reality is more dynamic. You forgot that currently there is not much incentive to cheat, so small amount of cheaters won't have any impact. If the responsibility is pushed down to consumers, there will be all kind of cheats popping up and would spread out wider than it is now.
Anyway, the ultimate reason for doing this at manufacturer level may come down to how easy the situation could be controlled (by you know who).