I have noticed other non-intel "not being slow enough" when deturboed but not sure what's going on. I thought at first, the turbo function of the board was failing to deactivate L1 like it did on the intel, but cachechk said it did. Last one I was messing with was my UMC U5S, that's like ~1.5x faster than an i486sx normally but over 4x faster in turbo mode. But thinking about a 386 class board, it probably has no idea it's supposed to turn an L1 on chip cache off. Started me wondering if there's any other mode change. For instance on that board 386 apps no longer worked right on intel in deturbo, but I was able to launch one or two on the U5 in deturbo.
I will have to investigate behaviour on my TI486SLCs when I get a PT-319A going.
Turbo modes in general seem all over the place board to board. Some seem to try for a 5170 8mhz compatibility speed, some still trying for 4.77ish XT on AT class.
Unicorn herding operations are proceeding, but all the totes of hens teeth and barrels of rocking horse poop give them plenty of hiding spots.