The formal testing was done by EVE on their own cells, but the results are widely agreed to be applicable to any other thin-wall rectangular cells. Compression keeps the sidewalls from expanding and contracting as much. In the test, compressed cells (around 12-14 PSI) had DOUBLE the lifespan of the uncompressed cell in the test - all subjected to the same "charge/discharge/recharge/repeat" regimen. IIRC, it was roughly 5000 cycles versus about 2500. They discharged them pretty deep.
Here's a pretty good photo of the concept, in which I used 4 compression springs to push steel plates into the large "faces" of some LFP cells. The small sides , and the top/bottom, do not need to be compressed separately to achieve the benefit (and the EVE test pressed only along the "big" faces of the cells.) Although I have used steel plates on my two "smaller" batteries, most people use 3/4" plywood. Cheaper and easier to work with.
For a plywood example, see
https://diysolarforum.com/threads/pr...98/post-430244
One of my own battery packs (before adding a heater pad and wood "casing") is show in the attached photo.