Most mixers will have a "low cut" that basically works it's way down from around 80 hz. I love Mackie's owner's manual -- says something like "Always use the low cut on all your tracks . . . except for bass guitar, kick drum, and recordings of earthquakes." Always get a kick out of that one.
Anyway, a lot of mics, like the sm57 for example, will naturally start rolling off at about 200 hz. It's more of a gradual rollof, mind you, as opposed to the steep "falling off a cliff" cutoff point you'd normally get with the bass cut on a mixer.
The difference between the two is that the bass cut on a mixer just clears up any of the really low and subsonic stuff. Rumbles, tubles, and that sort of thing.
A bass rolloff, be it a switch or a natural feature of the mic, works more as a musical effect. One of the reasons it works so well on the snare drum or close-micing of guitar amps. The bulk of the snare's energy is usually around 200 hz, so it's a good idea to roll off anything below that point, which the 57 does naturally. Similarly, 200 hz is a good point to start rolling off on an electric guitar track, so as to free up some breathing room for the bass guitar, snare, etc.
Hence, one of the true values of the 57 -- the tracks it records require much less fiddling with after the fact.
As far as why this would be valuable over just eq-ing it after the fact, think of it this way: When I record a track on my computer, for fidelity purposes, I want it to take up as much of the 24 bits that I have available. Suppose a good portion of those 24 bits contain unnecessary bass content which I plan on cutting out.
Now I'm not going to pretend to be an expert on digital recording, and I could be totally wrong on this, so if anyone wants to chime in and correct me please do so. But it just seems more logical to me that it would be a more efficient use of my available bits and/or magnetic particles (if I'm recording analog) to use as much of them as possible on audio information I'm planning to use? As opposed to recording, say, 20 bits worth of audio and then shaving off 1/4 of what those bits have recorded (mostly unnecessary bass content) ? ?