Just dug out my old Eyeline Direct newsletters if anyone still remembers those. The May 1999 issue gave us chapter and verse on the mobo timing issues and what Commodore and their ilk did and didn't do about them. Here's an overview.
The main UK selling board was the 1.D.1 and this was pretty well fixed during manufacture. A few early, pre-main-production revisions, notably 1.A and 1.B were also sold and these also had timing workarounds fitted.
Commodore planned to correct the Budgie and Gayle faults and went ahead with revisions 1.D.4 and 2.B which didn't have provision for the fixes. They planned to fit updated chips, but by then, the company was sinking and these boards went on sale with no fix available. So,fitting fast accelerators, fast IDE controllers and clock port accessories to these may cause timing issues. Similar boards were manufactured by Amiga Technologies/Escom starting in 1995, but early sales (15-20% of those sold) didn't get fixed either.
Many of these boards were fixed by Eyetech to bring 1.D.4 and 2.B boards up to the original 1.D.1 standard, but some of them have 'insoluble' issues. These are found when using a Zorro 3 bus which requires the Budgie revision - and never happened.
Many people run 1.D.4 boards without issues. My old tower has one and has lots of heavy stuff on it. There's an '060, a Power Flyer, a Catweasel, Silver Surfer, etc. The only problem I had was when trying to run both a PicassoIV and a VooDoo5 together on the Elbox Z4 bus. Elbox confirmed it was timing problems after a lengthy diagnostic session by email. Other than that my 1.D.4's been faultless.
JaX