I'd like to keep it running with something like a clown fish so I can detect if corals have any fish diseases such as ich.
Just to put this out there - if the corals have ich tomonts on them, then having a fish in the tank will give the ich a host to feed on and enable it to reproduce, and you won't be able to treat the ich infection without killing the coral.
Personally, I'd just QT the corals without a fish for 60-76 days to allow time for any fish diseases like ich to die off (no fish to parasitize = no more ich).
Seems like keeping a fish would also be a good way to test new fish coming in?
Not necessarily - it largely depends on the stress levels of the infected fish. Many infected fish don't show symptoms until they get stressed, then have an outbreak; this is part of why fish can be seemingly healthy for months or even years, then suddenly have an outbreak.
Anyway, if the new fish are stressed or stress out the original fish (and are in the tank long enough for a disease to reproduce to visible levels on that original fish), then you may see the disease symptoms, but if they don't get too stressed, you likely wouldn't see any symptoms.
This is one of the potential flaws of observational quarantine, and it's part of why the recommended QT protocol involves prophylactic treatment (i.e. treating for disease before we see any symptoms).
As long as the QT doesn't have corals or inverts in it, though, you could treat any diseases that show up in it.