Blood 2 was, I think, a victim of its shitty engine and the jump to 3D and, I think, being rushed.
I wouldn't call Heretic or Hexen or Strife obscure except insofar as people don't play a lot of them these days. Strife was significantly less popular than Heretic or Hexen or Doom, but everyone who played Doom knew about those three, even if they hadn't played them. I'm going to go out on a limb and say that Strife was probably the first FPS with a branching plot. I thought the late-game really fell apart gameplay-wise but it was neat to see how the little micro-plot developed itself. The sigil was a huge fucking disappointment. You can see they wanted a superweapon that you'd use at a cost when you were in deep shit, but I always thought it ended up being a shitty gimmick. The mostly urban-industrial-gothic levels were kind of uninspired, too, compared to something like Thief--but then again Thief came out a lot later, so that's really not a fair comparison.
Hexen gets real repetitive after a while since every class only has four weapons. Even with magic items, that just doesn't cut it for the length of the episodes. I have the whack-whack-WHACK of the Fighter's first weapon burned into my head, though, and I can't think about that game without hearing it.
My point was that if Unreal 1's multiplayer failure hadn't occurred, then UT and its successors could have been bigger than Quake's multiplayer scene. Don't get me wrong. Unreal Tournament was hugely successful (and deserved to be) but it and UT2003 (and their TCs) always felt like second fiddles to Quake and its TCs. It's kind of an out-there suggestion but I suggest that Unreal was so terrible in multiplayer that I think a lot of young players just abandoned the series, and a lot of them didn't look back when UT rolled around. I could be completely wrong. I agree that UT and UT2003 stayed relevant for a long time, but there were some damn good TCs for UT and UT2003 that should've had a lot more players than they did.