Depth of field is governed by three things: focal length of the lens, distance to subject and aperture. Aperture is the one variable that your on-camera flash can potentially help with. Potentially, because as was noted, a lens (especially one the size of the 105) can cast a shadow in some circumstances. This does not seem to be the case with the AF Micro 105mm 2.8D on the D90. I gave the light a little advantage by using a Gary Fong Puffer to act as a diffuser (it robs about 1 stop) but also to move the illumination source about an inch forward along the lens.
I've posted two shots of a single Black-eyed Susan bloom taken from one setup on a tripod with the lens at nearly maximum extension - it's within about 1/10 inch of subject to sensor minimum distance - the best I could do with grippy tripod feet on a rough floor. I'm looking at the bloom as I write this: the greeny center measures about 3/8" (9 mm) across and is, roughly, a hemisphere. The thicket of stamens, etc add about 1/8" (3mm) to the radius. The whole assembly is about 5/8" (16mm) across. They project about 1/4" (6.5 mm) from the petals.
I only focussed once, before the first exposure. The first is 1/200, f8, iso 160, flash (manual, 1/16 power). The second is 1/200, f57, iso 400, flash (manual, full power) - this had to have its exposure increased by 1/3 stop to bring the histograms within spitting distance. The difference in perceived detail is substantial, whether you like the quality of the detail or not.
Except for the one exposure adjustment, downsampling and identical export sharpening both images are out of camera. So, the short answer vis-a-vis the on-camera flash would seem to be Yes. Thanks for provoking me to do the experiment. I'll admit to being surprised, perhaps even delighted.