You'll get better understanding of the topic from e.g. Characterization of the material properties of two FR4 printed circuit board laminates by Haugan and Dalsjø; the full PDF article is open access.
Sorry, but like many other reports, this is about the 'glass transition phase' at high temperatures to research the effect of the soldering and the effect of high temperatures on board stiffness. It doesn't provide any information about effects of putting local compression stress on FR4. Also the fact that the report was written as part of a summer internship makes it just another collection of datapoints.
Like I wrote earlier, I cannot find any peer reviewed articles about mounting stresses on FR4 or similar materials
for different styles of mounting holes. For someone who does not even recognize the difference between elastic and viscoelastic materials, that PDF is a good starter on how these things are measured, giving you a better understanding of the topic at hand, instead of picking individual sentences or images at random that you think bolsters your argument.
If you are hoping for a good, possibly peer-reviewed article on FR4 mounting stresses,
there do not seem to be any; only "anecdotal" evidence from PCB designers having done stuff for decades, and "understanding" based on materials physics (viscoelasticity) and mechanical engineering (controlling and localizing strain under various stresses using physical structures, i.e. shape).
The next best thing, in my opinion, would be a Finite Element Analysis with a few different forces, both on the mounting hole itself (corresponding to forces in the PCB plane) and thus with multiple different hole models, and on the PCB (perpendicular) with different holes symmetrically placed around the stress point in typical rectangular configuration. This yields images where the resulting strain is easy to see; compare to
this one[/ULR].
(That [URL=https://www.researchgate.net/publication/337968594_Analysis_of_Printed_Circuit_Boards_strains_using_finite_element_analysis_and_digital_image_correlation]publication is open access, comparing digital image analysis of real boards to finite element simulations. So even that one is just an example of the
type of such investigation, and does not answer the question at hand.)
A set of boards with strain rosettes and different mounting hole styles for measuring the actual deformation would be easy to manufacture, but even with minimum trace widths, you'd need some high-resolution test equipment to measure. You'd also want to have a good microscope to examine the results of a crush test, say with thin angular (not perpendicular nor tangent to the hole) traces near the screw head just to make observing the effects easier.
Using minimally thin test traces in inner layers close enough (too close!) to the mounting holes should show breakage at high enough deformations, but I suspect one one would have to test batches of boards, dozens of mounting holes of each type, to get any kind of meaningful results.
Even if
I did such a test and posted the results here, most of you would not believe me anyway; and it would be utterly wasted effort for everyone.
Thus, the only real possibility is for someone else to investigate and report.
I could be wrong here, even if I don't think so. My own opinion is based on relatively simple physics I've explained already, but seems to be countered with "I feel you're wrong" -type truthiness without any basis, which annoys me to no end. I'm done arguing from simple physics principles against vague gut feelings.