This has been a very illuminating discussion so far, and I've enjoyed reading all the replies. Like many others I have my "full realism" days and my "hit ctrl-e and go" days. I find the challenge of building realistic systems functionality to be fun and satisfying, but as a user I often find I just don't have the free time available to learn all those procedures for other planes. I totally agree with Rob though that the feedback and control "feel" of the plane is paramount, that its handling in the sim is a more essential part of suspension of disbelief than whether or not every single switch does exactly what it should or whether parts should fail in very specific ways. Framerate of course is a big part of that since we lack any ability to physically feel the forces at work on the airframe. Granted there's certainly some crossover too.
When dealing with pure systems complexity in areas that do not directly affect the handling it's a little easier I guess, you can build two versions much as A2A does, to satisfy both camps, but when it comes to flight dynamics things get a bit trickier and more subjective I think. For instance what do you do when the plane you're modeling has certain very basic real world handling qualities that at best could be considered mildly irritating to a sim pilot (they are in real life too), or at worst be downright frustrating and off putting? Do you make the handling and feel as realistic as possible and possibly suffer complaints from casual users, or dumb it down a bit to make it more usable and perhaps annoy the hardcore ones? I had to deal with those issues in certain places on the L-39 (thankfully not too many), but the MiG I'm working on makes that plane's handling quirks seem positively mild by comparison.
I guess again you can make two versions, but where do you draw that imaginary line?
The issue of making the flight dynamics meet the published specs is interesting too, since those are almost always derived from flights performed under ideal conditions and with a very experienced test pilot at the helm, and then of course add a little manufacturer "hype" padding.
An example is jets with a thrust to weight ratio of greater than 1:1. Yes those jets may have it, but the part they don't mention is that this power is only available at relatively high airspeed where the engines are getting their maximum efficiency. Standing still on the ground they simply don't have it. Always fun food for thought, for me anyway.
-Mike
