When is it appropriate to enquire about the status of submission of revisions?
At any stage of the publication process, you should feel free to check in whenever you think the referees / editors / typesetters / whoever have taken a reasonable amount of time to do the job, and you are beginning to worry that your paper is not being processed in a timely way. I don't think that is any fixed amount of time: in particular six months was suggested by one pure mathematician for papers in pure mathematics. It happens that I also use six months as a rough guideline (I am also a pure mathematician), but I will adjust in either direction if it seems reasonable. For instance, if the paper is so short and easy that a qualified referee could do it one sitting then six months doesn't make any sense. If the paper is 50 pages long and hard even for me to read, six months doesn't sound like enough time. Of course what I think is an appropriate amount of time is a subjective judgment but informed by my own refereeing experiences.
Would you use the same rule of thumb that you use for initial submissions (e.g., 6 months or so)? Or is it reasonable to enquire about the status of the manuscript sooner?
If it seems clear to you that refereeing the revised version should take much less time than the original version then sure, inquire (or "enquire," depending upon where in the anglophone world you are) sooner. In my experience, most referees do turn around minor revisions much faster than the original report. The flip side of this is that the editor will of course try to get the same referee to do the revisions, and the referee might have become busier in the meantime. Anyway, all you can do is ask.
Alternatively, would you be less inclined to enquire about the status because there is more to lose by irritating the editor?
I am feeling a bit jaded about the publication process right now -- the quality of service that an author can expect from their submission experience varies so wildly as to seem like a fairness issue -- but I am not so jaded to think that editors would allow their feelings of "irritation" at this (completely appropriate and expected) authorial behavior to influence their processing of the paper. I have seen a lot of weird stuff happen over the years, but I've never seen that. However, I've witnessed many, many instances of authors who don't stand up for themselves and get victimized because of it. So I think it is much sounder strategy to "irritate" editors more rather than less often. Either that or cultivate a kind of inner peace that allows these kind of delays to, truly, not bother you. For that I think it helps to do a fair amount of refereeing of one's own!