One of the most uniquely important features of paleoart is that it is capable of generating testable predictions that future fossil findings can support or reject. My favorite example of this phenomenon is captured by a 1915 drawing of "Tetrapteryx" by American naturalist William Beebe, who felt that dinobirds must have passed through a four-winged stage in their march towards powered flight. He was informed by his studies of Archaeopteryx and by the long leg feathers found on modern bird embryos, which develop briefly in the egg and then disappear before the bird is hatched. This hitherto undescribed creature, Beebe wrote, would have had long pennaceous leg feathers that it used as a flight surface in gliding from tree to tree. Almost a hundred years after Beebe first predicted such an animal, the animal was discovered sleeping beneath the stone in northern China: Microraptor gui.
No other area of scientific illustration is able to do this as effectively as paleoart. A paleontologist can make a prediction in a manuscript that a few hundred people might read, at best, but a paleoartist can draw a prediction that becomes a visible and salient hypothesis that scientists and laypeople alike can evaluate and which may become widely associated with that specific dinosaur until it is disproven.
I would like to make a prediction about a future fossil of Microraptor itself, based on the same principles I learned from Beebe.

This newest study is another piece in the puzzle of Microraptor's ecology - a puzzle which is looking more complete as time goes by. With over 300 undescribed specimens, at least three perserved meals (all different), a color study, conflicting scleral ring studies, and a myriad of biomechanical research, we now know a tremendous about this animal (at least for being dead 120 million years).
But can we extract a reasonable approximation of its diet and lifestyle from the available information? To answer that, one option is to look towards modern birds. But first, let's look at what we do know.