I'm interested in behavioral research and when I look at top schools related to my interest (Illinois, Indiana, Wisconsin, Emory, etc etc.) most of the students have MAccs, MBAs, or strictly BS. I do fully intend to utilize all possible electives in the MAcc by taking math related courses regardless however. BUT, I certainly don't want to be doing research where I'm using real analysis and topology anyway -- fortunately most programs I've seen with significant behavioral faculty offer tracks with mostly stats courses instead of the PhD Microecon.
I didn't know about your behavioral interest. In any case, I don't think many people are doing research that uses real analysis and topology -- I sure don't. But the point is that signaling models are widely taught in Economics departments and b-schools and as such, many (admissions) people have come to believe in them: If you can get through the topology crap, you can probably get through anything that's more closely accounting-oriented at the PhD level. The other stuff on my list (econometrics, microecon, advanced calculus) will help you for sure in basically any PhD program.
That said, you also need to be careful when you look at current PhD students profiles. At Indiana, most of those with MACCs come from a professional background, where they spent 2-5 years at an accounting firm or a regulatory agency between their MACC and PhD. Those without such work experience, whether coming straight from a BS or a MACC, either are clearly standout applicants (such as the Nat'l Merit Scholarship guy) or went through some kind of pre-PhD preparation.
As an example for pre-PhD preparation, see the last guy on the list of current students (Zimbelman), who went through the MACC pre-PhD track at BYU. Some info on that track is found at http://marriottschool.byu.edu/soa/newsletter/PDFs/SOAIS_NewsLetterW05.pdf
. Those pre-PhD courses at BYU are clearly what I was talking about. Frankly (disclaimer: opinion here), if you look at the leading faculty in the behavioral field, I don't think you'll find many who lack any kind of formal training in microeconomics, because microecon is the open door that leads to game theory, and behavioral research is basically all about finding holes in the game theory stuff.