So from IBM's POV, they could
support Linux - which by then had already been ported to the VM/370
and there was already talk of porting it to the later mainframe
iterations. I don't think anybody was even thinking of porting any of
the *BSD to IBM mainframes till much later, am I right?
This is not how I remember it going down.
There was an external-to-IBM "Bigfoot" port to S/390 (not S/370) that IBM was ignoring until it got alarmingly close to booting, and then all of a sudden there was an IBM port to S/390. Clearly (well, *I* thought it was clear) they'd had a skunkworks project for some time and Bigfoot forced their hand. (Unix v7 *did* run on S/370, and resurrecting that is one of my hobby projects that hasn't really gotten off the ground).
I was the system administrator of the first publicly-accessible Linux-on-S/390 machine--penguinvm.princeton.edu--and indeed in the late 90s I and my mentor David Boyes met with some pretty high-level people at IBM to advise them how we thought they should proceed. They seemed to take much of our advice, but then again I don't think we said anything very crazy. (At the time, and for years thereafter, I was with Sine Nomine Associates. They're still around.)
I also later managed the port of OpenSolaris to zSeries, which, if IBM had bought Sun rather than Oracle, would have made my life very different. Neale Ferguson did most of the heavy lifting on that port, but I did a lot of the tool porting and wrote a disk driver. Alas, IBM tightened the screws a little too far and apparently didn't know that Sun had an offer from Oracle in its back pocket.
But back to the S/390 port--I went to a Linux conference in Atlanta in the late 90s ('99, I think) to speak about Linux on S390/Z, and I actually went by the NetBSD booth to say, "hey, I can maybe hook you guys up with a development virtual machine," and what I got was an earful about "your so-called portability" from someone who was clearly much more invested in hating Linux than in, you know, saying, "wow, OK, I realize you're not offering me cycles on a super-awesome machine, but, yeah, it's not nothing, cool, here's who you should talk to if you're interested in getting a port going."
So I don't think you can lay all the blame on BSD inaction on Linux, is all I'm saying. By '99, I think it was, maybe if NetBSD, which already had its reputation for spectacular portability, hadn't staffed its booth with a jackass still trying to fight the Unix Wars, that story might have turned out differently.
Adam