If you think about it in terms of the Canonical Ensemble the regions of states that correspond to "a brain and body" are obviously a small subset of the regions that correspond to "a brain". The incompressability of phase space means that this translates directly to the world being less likely to wander into those smaller regions via random fluctuations in direct proportion to their size.
In the end, the second law of thermodynamics is only a (very strong) statistical likelihood. If you mix some salt water and fresh water, then wait long enough you'll eventually find all the salt in one side of the glass (in time proportional to 2^N where N is the number of salt ions). Its not something written into the fabric of the universe directly, but a consequence of more fundamental laws and if you take Statistical Mechanics in college you'll get to re-derive it and see why it works that way. In fact, I'd really recommend taking a Stat Mech class to anyone considering it, it was probably the most interesting class I took in college.
Anyway what is the point of this (Boltzman brain paradox)? (or point that was supposed to be?)
The horror of the Boltzman Brain idea is that the vast majority of the conscious beings that will ever exist will be created in environments that cannot support them, and will promptly go insane and/or die.
Universe filled with many brains fluctuation can be more simple than brain fluctuation. Fluctuations may also happen in environment that doesn't even have atoms that are needed for brain , but will have atoms that are needed for universe. Maybe this atoms needed for brain can only be created by universe fluctuation.
If we assume all this unknowns and create our imaginary universe yes there can be more brains in it than universes.