If you use the MS effect only on playback, monitoring during recording will sound rather odd (the whole thing will sound right-sided as memory serves me). But you can then dematrix the MS pair to normal stereo in post-production.
If you use the MS effect only on record (you certainly do not want to use it on both!) monitoring will sound normal, but the MS pair will already be mixed to normal stereo. However, that doesn't actually matter, as in post production you can use a vst effect like "Voxengo MSED" to turn normal stereo back to MS, alter the width, then turn it back to normal stereo in a very fuss-free manner.
So personally I prefer that route. YMMV.
It's a pity that effects on replay can't be applied during monitoring too without being recorded.