I agree, Calibrator, one should never forget the audience (neither the author nor the critic!) These games weren't made for the average Joe. They were made for highly literate, college educated, computer wizards, many of whom had extensive training in advanced mathematics and engineering. I wouldn't be surprised at all if Airfight was designed by a real aviator. That said, to be fair, the games do have pretty extensive help files available if one cares to read them. Most of the information is there if you really want to take the time.

Ultimately, though, I think you have to always think about the common person. If you ever start thinking in terms of "this is only intended for specialists," your actual influence will be severely limited. In my opinion, there is nothing so complex that can't be broken down into enough steps for the common person to climb them. You just have to make sure you aren't taking things for granted, going over someone's head or moving too quickly. In my experience, someone who claims that X can't be explained because it's too complex doesn't have a firm grasp of it in the first place. Anyone who really knows a subject can easily teach it to anyone who sincerely wants to learn it, given the time and motivation to do it.

Coming back to the topic, though, you had mentioned Ultima clones. I think that's a great example. If you want to make a very successful Ultima clone, you should start by thinking about gamers who have never played Ultima or never even a CRPG, period. Think about what it is about the game (getting down to the very core) that makes it appealing. Then, once you've figured that out, you're basically just building a bridge between yourself and what you find enjoyable and the average gamer and what he or she already knows how to do. The game itself will be greatly improved if you have taken the time to fully understand it; not just the code, mind you, but what makes that gameplay worth coding in the first place.