Of course, it is a matter of taste.
You're correct that a chorus creates multiple “voices”, but actually that's a good reason to put it after distortion! For distortion is nonlinear, which means that if you put in a combination of multiple voices, the result will be different from when you put each voice individually through distortion and mix them afterwards. In particular, you get intermodulation artifacts, which is what's often considered a muddy sound. However, intermodulation isn't necessarily bad – for instance, powerchords wouldn't sound nearly as powerful without it! Likewise, a gentle chorus before not-too-brutal distortion gives a thickened-up sound without too much artifacts. Also, the distortion will somewhat “obscure” the character of the chorus itself, so it won't sound too obvious, artificial, or “80's”.
OTOH, a chorus is actually linear (or approximately linear), so technically it really makes sense to put it after distortion: chorus after distortion actually gives almost the same result as you'd get by putting each individual chorus voice through a seperate distortion unit.
But again, “technically correct” doesn't means it's musically the best thing. I personally tend to default to chorus-last order, but I like occasionally trying how the chorus-first sounds in a particular context.