I like this idea.
IRL you only need one engine failure at the wrong time to kill you. This is quite a powerful argument for conservatism. In the sim, we've got a refly button and therefore are likely to fly rather differently even if the other aspects of the model are perfect.
Basically you'd have from sunset to sunrise to get a squadron of day fighters maintained, and in the height of summer you'd need a lot of manpower to put in the man-hours per flying-hour required in the short window of darkness available. If you'd didn't get the job done then you wouldn't be able to scramble a full squadron.
It's very unlikely that corners would be cut until the airforce in question was really falling apart, because it's better to lose a sortie than to lose an aeroplane & pilot.
Therefore, the real driver behind the increased risk of failure associated with pushing airframes and engines hard would be that life consumption might exceed that assumed in the maintenance schedule.
However, in the wider scheme of things, the maintenance schedule was a constantly moving target, adjusted in the light of operational experience. If the accident rate went up, the maintenance intervals for suspect components would be reduced.
So, unlucky pilots would be unlucky, but at a force level the real impact would be upon the sortie generation rate achievable with the available manpower, which would vary gradually, over a timescale of months rather than days.
I therefore think that the only way to encourage reasonable flying with a fully realistic model outside of a long-term campaign environment is to dock points for rough treatment of airframe and engine, because otherwise you'd end up having to deliberately make the models unrealistic in some way.
Having said that, if you're going to tweak the model for gameplay-friendly wear/failure behaviour, then a wear modifier as suggested above is a good way of doing it, especially if it's connected to the external weathering model. Pilots could then select their individual airframe in the armament screen rather than just the type and loadout, which would add extra depth to the simulation.
Pilots scoring lots of kills could get extra maintenance (and possibly the addition of an ashtray and telescopic sight to their 109's cockpit model if they're extremely successful
)