Convention says autowah and similar goes first (something intended to react to pick dynamics won't work well with a compressed signal for obvious reasons).
Compressor generally next. Though if the guitar signal is very clean (in terms of no buzz) and the compressor doesn't add much noise of its own sometimes they work well after distortion/overdrive (assuming the distortion/overdrive doesn't add much noise itself, of course).
Wah after, before or in between compression and distortion depending on preference (it gives different noisefloors and different sounds).
Chorus, flanger, phasers, tremolo etc. after that (though again, sometimes running these into distortion can work well).
Then delay/echo, reverb last.
Unless you're Leo Fender designing an amp, when you put the tremolo* (or "vibrato" as he called it) after the reverb, which gives a much more marked tremolo effect as the reverb doesn't blur the volume dips.
In the real world though, it's a matter of experimenting until you hit on a sequence that works well for you.
As for effects loops, yes there's a difference. If your distortion is coming from the amp's preamp stage then the gear in the loop will affect that sound. If the effects are pre-distortion (i.e. in front of a distorting amp) the distortion works on and reacts with what the effects have already done. Both setups are perfectly valid depending on what you want to achieve. FX loops generally work at line level, not the lower instrument level that the guitar produces, by the way.
Reverb built into an amp is generally in between the pre and power amp stages, so is in effect already in a "loop".
*tremolo is volume fluctuation. Vibrato is pitch fluctuation. Fender got this backwards for some reason then stuck with the error. So a Strat's "tremolo" bridge is really a vibrato bridge, and Fender amp's "vibrato" is actually tremolo).