Emulating the machines you list in an FPGA only really makes sense if you put the microcode in block RAM, until recently the devices at prices most people could

Emulating the machines you list in an FPGA only really makes sense if you put the microcode in block RAM, until recently the devices at prices most people could afford didn't have enough block RAM to do this. In the meantime, most input devices have moved to USB which is a lot harder to interface, someone could write a USB stack in Lisp for MIT/LMI machines or for the Xerox systems but it is a big task. The Tang Nano architecture of using a separate conventional CPU for IO seems like cheating to me though I suppose it does match how Symbolics machines worked.