Do any of the really early versions of X (and W) still exist?
Not that I have found. I looked a couple a years ago. FWIW: I would be surprised if any of it was in CLU, because the first version of X was a redo from synchronous W protocol to the async one Bob did. If I remember it correctly ... W as for the V Kernel on the Stanford (68K based) Sun Terminal, that Paul Asente and Brian Reid did at Stanford (in C). The Paul and Chris Kent at DEC (either WRL or SRC I foirget which) took it and made it work on a VS100 and microvax. That version when to MIT via the Athena connection and Bob rewrote the back-end creating the X from W.
Shortly after that Gettys brought the Microvax X version out to me in Westford where, he, Jack Burness and I got it running on a 68K based Masscomp over a weekend fairly soon after Jack his first version of his new BITBLT primitive working - which until then we lacked (it was pretty cool how quick it came up).
Clem