> These two kinds of objects have comparable numbers of stars and are roughly the same size (though dwarf spheroidals are admittedly larger by about an order of

> These two kinds of objects have comparable numbers of stars and are roughly the same size (though dwarf spheroidals are admittedly larger by about an order of magnitude). This discrepancy can be explained with cold WIMPs [2], but it's hard to understand how some MOND-like process would produce such incredibly different dynamics between the two kinds of galaxies. I think the order of magnitude size difference would be important in the possible origin differences in a MOND-ian universe. If globular clusters formed in a higher-pressure environment (within a high-z galaxy or merger) they could be more compact and stay in the Newtonian regime. And if dwarf spheroidals formed in lower-pressure environments that might be the origin of larger sizes, putting them in the MOND acceleration regime. But, that's just speculation. And to emphasize your statement that "the evidence for dark matter is much more than galaxy rotation curves these days", even MOND-ian models require some form of "dark" matter to explain cluster dynamics.