It was not popular with CS Research, and we were not popular with them. We were using VAXes, which the 3B series were attempting to compete against. The VAX was not exactly graceful, architecturally, but the 3B series was clumsier and less cost-effective. We weren't interested, despite frequently applied pressure.
And, although a different topic, there was the way the commercialization of the Blit forced the 68000 to be replaced by the BELLMAC-32 by essentially the same people, or at least the same boss (Scanlon), which was a poor decision on every dimension. The idea was to get BELLMACs out there to drive up production, but the chip was far less suitable, and each one cost about what a full Blit with a 68000 instead would have. A loss leader, bad financially and bad technically.
On the other hand, as I resullt of I did port an OS and other software to the BELLMAC-32 as a result of this work, and learned how badly it did things like memory management and the interrupt vector setup, and how buggy it was.
So there were not many warm feelings between 1127 and the computer division.
-rob