fwiw, i have run two 671's and the clocks were perfectly in sync. just had to line them up once. so, its probably the case with two r09's.
FWIW too, I've used two different models of Hi-MD reorders on a job and was amazed at how they held in sync throughout a classical string quartet performance.
I'd attribute that to either (1) a relatively short performance, (2) not looking looking for the drift at a sufficiently zoomed-in level of detail, or (3) pure luck. You two probably just had very little drift, to the extent that it was not audible to your ears at the beginning vs. the end of the show.
The assumption that two of the same type of recorders' clocks will drift less than the clock of two different kinds of recorders is probably a safe one; however, it is a bad assumption that the different clocks of two of the same type of recorders will not drift. My experience has been that there will always be at least some drift between the clock of two different recorders (of same or different types), and I usually see it on the DAW as soon as 5-10 minutes into the show when zoomed in sufficiently (at ~ the detail of milisecs or even samples level of resolution).
QFT.
As someone who's synced all sorts of crap, audio, video, different recorders, different videocams, etc., the only thing that's "safe" to assume is that you will have drift for anything more than an hour. Whether it's audible or not will depend of course, but don't assume you'll still have sync at the tail. Also, if you're not careful, you'll end up with crazy weird phasing somewhere in the set, but you might not notice it until after listening to the whole thing again.
FWIW, I usually sync the head, then check the tail for sync, and stretch (using a good editor of course) one of the sources if needed to get perfect sync throughout. Another approach is to try syncing at the midpoint of the wav file, that way if you are drifting, it'll only drift half as far at the two ends as it would if you synced just the head of the wav.