The fastest FPGA Arcade and Mist cores (based off the TG68) will likely never exceed 68040 performance
You might want to word that more carefully.
The 68K implementation running on the Mist/FPGAA of _today_ might stop at 040.
That same core can be re-targeted for, say, a Stratix 10 chip (we _are_ used to pay through the nose for our hw anyway) and it would probably be 2-5 times faster, mostly as-is. No, there are no FPGA designs out there with that chip on, but why should the Mist/FPGAA be the end of the evolutionary line?
