Signal-to-noise specifications for microphones are tough. The use of "A" weighting is inappropriate to begin with (it's designed for the way human hearing works in the ~40 dB SPL range), and it is almost always associated with time-averaged measurements which "smooth out" any impulse noise. Because of this, "A" weighted RMS values of different microphones can't be compared directly, unfortunately. Certainly a difference of 2 or 3 dB in the "A"-weighted RMS S/N specification between two different manufacturers tells you nothing that you can count on; the actual difference could be 0, or could be 2 or 3 dB in the opposite direction, or it could be a 6 dB difference or more in reality.
CCIR measurements (which are almost always quasi-peak, not RMS) will be noticeably higher--often by as much as 10-12 dB--but when accurately stated, they correspond much more reliably to what human ears will hear. The hitch is that most manufacturers don't list CCIR noise measurements precisely because they are higher, and even when they do, you don't know whether they're stating a typical value as actually measured in production, or perhaps an "aspirational" or "conceptual" specification.
The manufacturers that make the highest-quality microphones generally offer the most careful specifications, but the result is that the less careful manufacturers end up with specifications that resemble those of their betters. So it's a chicken-and-egg problem: You need to know who the highest-quality manufacturers are, in order to know how seriously to take the specifications that ought to tell you which manufacturers are offering the highest-quality microphones.