24 of my 36 Mercurian craters have been almost fully analyzed at this point, and so a plot akin to what is present in Neish et al. (2014) and Neish et al. (2017) has been derived for these craters:

Moon vs Venus vs Mercury (updated)

The number of “coincide” craters is small while the percentage of “~90*” and “>90*” craters is notably greater, as it is with Venus; however, there’s a large percentage of “Within 45*” craters — more like what is seen with the Moon.

It may be that the heavily cratered nature of Mercury’s surface is responsible for the “Within 45*” anomaly, but otherwise it appears the factor primarily controlling impact melt emplacement on Venus (direction of impact) is also predominant for Mercury.

This is where looking at rayed craters on Mercury becomes important, as the shape of the crater rays can give insight into the direction of impact.  Here are a couple obvious examples on Mercury (using the MESSENGER Colour Mosaic map):

This slideshow requires JavaScript.

Assuming a crater’s rays are asymmetric (the Xiao Zhao image illustrates this nicely), one can simply mark the region where those rays are minimal to imply a direction of impact.  Using Xiao Zhao as the example, the area where the fewest rays exist is roughly the NNE implying the projectile struck the surface from the NNE.

And then there are craters like Hokusai, that exhibit no such trait; all its rays are, more or less, symmetrical in their radial pattern.  Does this imply the impactor that made Hokusai struck ground roughly vertically?

The remaining 12 craters of my catalog need MDIS-NAC imagery to supplement the mosaic imagery already present — unfortunately, said imagery for these craters is not the best, so the MDIS-NAC is meant to clear things up enough to resolve the melt pools and/or crater rim and floor.  Once I can get Isis3 to cooperate with these image files, I can post another useful update!