After a rough survey of current and potential MUD admins out there, I have decided to shift my approach from a competitive stance (against what is out there) to a purely experimental stance (against nothing). I see little to no use in trying to make a system that is already experimental in nature conform to what existing and new admins would expect from a server. Thus, my work will not use potential overhead in immediate factoring of what technology to use and include in design and implementation, giving me a far more broad range of possibilities.
I have resigned in my attempts to get MySQL to play nice with the dlopen()'d modules that are the backbone of the MUD, and have switched to using BerkeleyDB. This move will greatly improve data access overhead and speeds, among other things.