James L. Carroll - if I understand what you are saying, you are suggesting that since information is lost in the process of nesting worlds, the exterior worlds contain more information.
Further, you are saying that an observer in the simulated world takes up information-space in the exterior world.
Further, you are saying that since simulated observers take up information-space in the exterior world, there are more observers "occupying the exterior world" than observers occupying the simulated ones.
That is true if by "occupying the exterior world", you mean "taking up resources in the exterior world". But that seems to miss the point of the simulation argument.
The simulation argument can be re-stated as the assertion that observers occupying multiple levels of simulation outnumber those occupying only one level.
We could make a similar argument with respect to the way our computer systems are built. Technologies are far more likely to employ several levels of processing than they are to employ only one level of processing.
When taking a survey of all the code in existence, I'm betting that most of it uses some form of layering, and very little exists only at the "metal".
Similarly with observers. It seems very unlikely that most of us exist only at the metal. If we have simulation technology available, most observers will end up occupying multiple levels of simulation.