Unfortunately there is no easy way to compare something like AmberMush to Artic either, it's like comparing finger-painting to football. The easy way to get to your conclusion is to compare similar games with relatively equal exposure. Checkers is more popular than chess. Hearts is more popular than bridge. It's not the conclusion I question but how you get there.

Of course the less is more principle isn't really going to help the OP, considering those expectations have already been set. When you're stuck with a particular game design, there's not a whole lot you can do short of a complete overhaul and rewrite to dramatically change the audience, make it a lot more appealing, or extend the average player lifetime. That's not to say that Diku-style games couldn't do with a massive interface facelift and polishing, but there doesn't seem to be much interest in that from those creating derivatives. They seem to prefer 'hey let's add more stuff' approach.

But to go further afield, I think someone using the less is more principle coupled with an onion or russian doll style unfolding game might be able to stake out a nice sweet spot in text muds.