I was asking more about the productivity benefits, versus any performance boost a "made for lisp" ISA might offer...

Productivity measured against what quality metric? For getting code written fast, Lisp is pretty good. For writing an operating system on which you can run a web server that will be competitive with nginx on Linux in terms of requests-per-second, probably not so much.

Maturity? Gets out of the way? Advanced documentation/learning material? Wealth of prior work solutions to common patterns? It's not so hard to guess what any developer's productivity is hinged on.

I asked because many people that used them called out the machines as being very productive...above and beyond lisp. Another comment here suggests it might have been CLIM.

Ah. A lot of that was because the LispMs had really good IDEs at a time when IDEs were all but unheard of. This was possible because it's a lot easier to write an IDE for Lisp than for C. But all that can be re-created at the application level (obviously because IDEs are applications). There's no need for a system-level Lisp to get that win.

In fact, Clozure Common Lisp is a perfect example. It provides (IMHO) 80-90% of the productivity advantages of a LispM when you run it on a Mac (because the IDE is Mac-specific).

I'm intrigued too, what's the 20-10% that's missing that was present in the LispMs?

Mainly the ability to write and debug system code with the same ease as you debug application code.