I put a third-party FYSETC Mini LCD32 in my Prusa Mini. It worked, then a firmware update broke it: the screen flashed white about every two seconds, on and on, unusable.
Prusa's own display is fine. The aftermarket one differs in one detail that turns out to matter.
A pin, not a timing problem
The ST7789V controller can run full or half duplex, selected in hardware by its IM3 pin. Prusa's board pulls it one way; the FYSETC board leaves it LOW, which puts the controller in half-duplex mode.
In that mode the display still answers when the firmware asks it to identify itself — but the reply comes back on SDA rather than SDO. The firmware is listening on the wrong wire, reads nothing meaningful, decides the display needs reinitialising, and does so. Two seconds later it asks again, gets the same non-answer, and reinitialises again. The white flash is the reinitialisation, repeating forever.
The fix
Stop asking. The firmware's periodic health check adds nothing on hardware that cannot answer it, so the patch comments out the body of the reinit-and-redraw routine — four statements. The screen stops flashing and works normally.
Not elegant. It is a local patch for a board Prusa never intended to support, and the right long-term fix belongs upstream in how that check handles half-duplex panels.
Credit where it is due
I did not discover the IM3 explanation. Prusa issues #3561 and #3277 had already established it, and the diagnosis here was reading code and datasheets against that, not instrumenting hardware. What I contributed was the working patch and a flashing guide, for the specific case of this display on this printer.