Another reason is that "red dwarf" is actually a very broadly defined term.
See this:
http://en.wikipedia.org/wiki/List_of_nearest_stars
Out of the 54 "stellar systems" within 5 pc of Sun, including the solar, 9 contain only brown dwarfs. This leaves 45 stellar systems, with 61 stars.
Of these 61 stars, just 15 are not red dwarfs (and of these, 4 are white dwarfs, leaving just 11 burning stars). So yes, 3:1... without white dwarfs 4:1.
But of these 46, the brightest is AX Microscopii, at +8,76 absolute magnitude. The dimmest is SCR1845-6357 which despite its magnitude +19,41 is recognized as red dwarf, not brown. Thus the brightest red dwarf is 10,65 magnitudes, or about 18 000 times, brighter than the dimmest one. And for comparison, the brightest star in the region, Sirius, is just magnitude +1,42, so about 7,34 magnitudes brighter than AX Microscopii - less than 1000 times as bright.
In terms of mass, AX Microscopii, at 0,60 solar masses, is about 8 times more massive than the dimmest red dwarf, yet only 3,4 times less massive than Sirius.
But while all stars so much lighter and dimmer than AX Microscopii are one spectral class M, the few brighter stars are divided in 4 different spectral classes: A, F, G, K.
Grouping the stars by first letter of spectral class:
A: 1 (Sirius A)
F: 1 (Procyon A)
G: 2 (Sun, Toliman A, tau Ceti)
K: 6
D: 4
M: 46
But see the first number of absolute magnitude:
1: 1 (Sirius A)
2: 1 (Procyon A)
3: -
4: 2 (Toliman A, Sun)
5: 2 (Toliman B, tau Ceti)
6: 2
7: 1 (61 Cygni A)
8: 3 (AX Microscopii, 8,76 is red dwarf, 61 Cygni B at 8,31 and Groombridgge 1618 at 8,16 are not)
9: 1
10: 8
11: 8 (1 white dwarf Pup, other 7 red)
12: 1 (white Procyon B)
13: 7 (1 white dwarf)
14: 6 (1 white dwarf)
15: 10
16: 5
17: 1 red dwarf and 1 brown dwarf
18: 1 red dwarf
19: 1 red dwarf and 1 brown dwarf
In the range 10...16, the red dwarfs are still rather numerous.