I'm not sure that the photos in the report are of sufficient resolution on their own to say much really. I'm trying to find a 1937 dictionary to see exactly what the definition of the word "crazing" was in 1937...I think we could then, with a good degree of confidence, take that definition to be what the learned authors meant in the report.
Meanwhile,
Per Ric -
The curvature was tested by placing the artifact against the cross-section view in the Lockheed engineering drawing
Note that the Plexiglas remnant measures approximately 4 ¾” in its longest dimension (see first photo below)
The Lockheed drawing defines the cabin window curvature using a straight base line marked with one inch stations and offset measurements to the outside window surface at each station.
http://tighar.org/Publications/TTracks/12_1/PartNumber.htmlThe copy of the Lockheed drawing shown in the second photo below, was used to demonstrate a curvature match with the artifact, but appears to have been significantly enlarged beyond actual size, evidently by approximately 150%, when compared directly to the known dimension of the artifact (approx. 4.75”).
Since curves are defined by their radii and changing the scale/enlargement of a drawing of a curve also changes its radius, meaningful curvature matching can’t be done by overlaying an actual curved object on a drawing of that object
unless the drawing is actual size so the curvature match between the fragment and drawing may need to be confirmed using a life size copy of the drawing.