> was there ever any crossover regarding UNIX folks applying their developments to other non-UNIX AT&T systems

Besides Sandy Fraser's long-term effort to advance digital communication (as distinct from digital transmission), there was TPC; see TUHS https://www.tuhs.org/pipermail/tuhs/2020-April/020802.html and other mentions of TPC in the TUHS archives.

Ken Thompson did considerable handholding for early adopters of Unix for applications within the Bell System, notably tracking automatic trouble reports from switching systems and managing the workflow of craftspeople in a wire center.

Bob Morris's intimate participation in a submarine signal-processing project that Bell Labs contracted to produce for the US Navy set him on a career path that led to becoming chief scientist at NSA's National Computer Security Center.

Gerard Holtzmann collaborated to instill model-checking in switching and transmission projects.

Andrew Hume spent much time with AT&T's call records.

Lorinda Cherry single-handedly automated the analysis of call centers' notes on customer contacts, This enabled detection of significant human-engineering and public-relations problems.

An important part of my role as a department head was to maintain contacts with development labs so that R and D were mutually aware of each other's problems and expertise. This encouraged consulting visits, internships, and occasionally extended collaboration or specific research projects as recounted above.

Doug