As F107 says, I think it's a function of room size. In a room that's not too big for the M22s there won't be much noticeable difference between 22s, 50s, 60s, or even 80s in terms of "ability to fill the room". BTW the same goes for M2s in a small enough room.
As the room gets bigger, you need bigger speakers (probably bigger driver area is what matters) or you start hitting the limits of what the speakers can do comfortably. In a 13x23 room M2s "hit the wall" when you try to play loudly -- I imagine M22s with twice the driver area would be noticeably better in that room.
Make the room bigger again, and I expect the M50s would really come into their own. If you measure the active area of the drivers (using the center of the surround) you get roughly 4" for the smaller drivers and 5.5" for the larger drivers -- 12.5 vs. 23 square inches of surface area respectively -- so the difference from M2 to M22 should be about the same as the difference from M22 to M50. Interestingly enough the M3 and M22 have about the same driver area so it probably is frequency response that accounts for one sounding larger than the other standalone.
I forget the term -- might be "dynamic compression" -- when you reach the point where cranking the volume up doesn't make the SPL go up as much any more, but it is noticeable and the M22s should hit that point before the M50s.
From the Axiom site -- "A sub-sat system... with the M22s... gives up only the ability to play at extremely loud levels in larger rooms to floor standing speakers...".
EDIT -- for ricwilli's room (11x19x8 ?) the M22s would be more than enough for HT (ie I don't think they would hit their limits at all), only concern is how loud "loud music" really means. For some people (self included) loud music can be REALLY loud
