At last, a discussion of polygons from jpl. We've seen them before, but not quite like this.

I support a dessication model. Polygons often from by contraction, like a cooling basalt or a drying evaporite. These are interesting, though.

A couple of things jump out while looking at these images in 3D.

Let's start with this one:

In my opinion, our view is perpendicular to bedding, that is "normal" to bedding, a birds eye view of a deposition surface. Only that tilted square of interesting texture is the top of the bed (you have to look at the stereo pairs, the area to the left is "deeper" than the area to the right). As to the interesting "ripples", they appear to me to be either "flute marks" or "load casts". Flute casts form when sediment is reworked by a current. Load cast form at the interface between different beds as the sediment pile is de-watered and/or vertically compressed by the accumulating sediment. But it usually happens at the interface between two different sediment types (mud and sand, for example) You will note that these are different than the polygonal cracks. I'll choose flute marks on a thick bed of sediment.

From another view:

We can imagine a very weak bedding running along the long axis of the rock, inclined down to the right. The top of the bed is the interesting "bedforms". Below that (to the left) the rock appears to be a massive variety (facies) of the hematite nodule bearing sulfate rock we've been following. Only with thicker bedding. Those joints (cracks) really do look like desication structures, but with a basal parting. If my bedding orientation observation is correct, then not only do the polygons go down from the surface of the bedding plane, but they also part along a plane that is parallel to the bedding. Ever seen mudcracks curve upward and almost completely detach from the surface ?

Another speculation is that the polygons in this view represent a drying surface at an angle to the orginal bedding surface. Thus the rock was moved while wet, then dried out.

Observations: we've seen polygons before, and outside the craters as well. I've speculated that the interesting linear structures outside the crater might map out as polygons but have not seen supporting orbital imagery to satisfy my conjecture. On the way to the heat shield, it might be good to map out some of the trenches to discover their geometry.

I'm still supporting a repeated wetting, deposition, and drying cycle that created a thick pile of sulfates that represent many, many such cycles. That happened long before the crater was formed. If the blueberries are lag deposits, then a whole lotta sulfate has eroded away before that impact event.

A metorite impact could cause a dewatering if there was water still in the sediment - that's what happens as liquifaction during an earthquake.