I don't know about everybody else, but yep, I consider subject vibration too.
Usually it's not an issue because the natural frequencies are high and there's quite a bit of damping, so any transient responses die out too quickly to notice. Once in a while something resonates enough to be a problem. I had to give up on one mounting scheme involving a 3-axis rotatable stage made with half a billiard ball sitting on 3 supports. Turned out, it resonated with the fan of my illuminator at certain brightness settings. The vibrations were noticeable, and worse, they were circular, so the blurs looked like lens aberrations or diffraction. That one took a while to figure out.
That case, by the way, illustrates that for some kinds of lighting, you really want the lights to be de-coupled from the camera and subject. Charlie Krebs has pointed out that my setup is less than ideal because the illuminator fan vibration can link to the subject stage. I agree completely, and as I wrote above, I've actually seen that happen on occasion. But given various other constraints, it has always seemed a bit more trouble to move the mount someplace else than to deal with the vibration some other way. If at some point it becomes more of a problem, or I see a great way to fix it, then of course I'll do that.
With the lacewing egg (here), I was more concerned than usual because of all that weight (the egg) stuck out on the end of that long skinny stalk. Turned out there was no problem at all, I suppose due again to a high resonant frequency and lots of damping from the silk.
An important issue in all this is where the vibration comes from.
In Oxford's shipboard case, it sounds like their camera system was internally quiet, so what they needed to do was lock camera and subject together, and isolate the combination as best they could from the outside world. High stiffness, high mass, and spongy supports carry the day in that case.
In my system, vibration comes primarily from three sources: 1) transient jerking of the mirror and shutter in the camera, 2) continuous oscillation from the illuminator fan, and 3) whatever the outside world does. In most cases, that seems to also be the order of importance: camera first, illuminator second, outside world third (now that the kids are gone...).
It is surprisingly difficult to mechanically damp the shutter movement. In my setup, the camera ends up being rigidly coupled to a hunk of machinery that weighs something like 25 pounds. My intuition says that should be plenty, but my intuition is wrong. If I place a finger gently on the milling table, I can definitely feel a tiny jerk when the shutter opens. The whole hunk of steel moves! With ordinary photography, this wouldn't matter at all, but at 2000 pixels per millimeter on the subject, what can be felt can also be seen.
It would help a lot, from a vibration standpoint, if I could isolate the camera from both the lens and subject. That's essentially what Charlie's micro setup does, suspending the camera over -- but not touching -- all the optics. Again, I haven't figured out any reasonable way to do that, given other constraints like needing to keep everything aligned while also changing extensions, swapping lenses, etc.
Keep those thoughts coming, Dave. At the very least, they make me think & write about why I do what I do. And I'm confident that one of these days I'm going to have a "Doh!" reaction and realize that I've been working the problem the hard way. That'll be a good day.