The paper references SPUR which ran Common Lisp, so they should have known about then current work on optimizing lexically scoped Lisp. They could even have lo

The paper references SPUR which ran Common Lisp, so they should have known about then current work on optimizing lexically scoped Lisp. They could even have looked at retargeting Franz Lisp at a register window machine to get some numbers.