Thanks guys - but to clarify -
The first image was 60 minutes, with agitation at 30 mins. - what some people are calling "semi-stand". I've found numerous posts (with scanned examples) where people use this with Rodinal successfully, many of them considering it their favorite process. I've used it on 35 and gotten some nice results. I felt it looked too flat. (And as I mentioned above, I felt the streaking came from lack of agitation). (And for 35, I like how the semi-stand can work to tame contrast). I'm quite willing to try about anything (HP5+ in Dektol paper developer is really kind of cool, for instance).
As I mentioned in the OP, I went to a more "recommended' or standard method - the 2nd two images were done as follows:
Rodinal 50+1, 20c;
11 minutes (film rated 400);
Agitation 1st 30 secs; 5 seconds every minute; and final 30 seconds. 8 seconds to pour.
My tank can take 2 35's or 1 120 reel; I need to make a spacer to keep the 120 reel at the bottom of the tank (so it won't pump up & down). Instead of inversions, I swirled the tank - sort of an almost-complete inversion. That's something I've done for years (when the lid of my older tank leaked too much to fully invert it). I do this for 5 seconds every minute. Though the reel doesn't go fully upside-down, I feel like it's getting plenty of agitation. But no air hits the film this way, as it would with an almost-full tank being fully inverted.
My main question was more of "can processing errors produce something that looks more like a ghost image than a smear". I can't think of a way that it could.
I'll do some tests for camera shake - the last I used that RB was for fashion catalogue work 10-15 years ago which was strobes and higher shutter speeds. I just re-foamed the whole thing so it's properly damped... but that is one big-ass mirror...
And wondering still - can paper be expected to reproduce all the tonality of a neg? I assume paper's tonal range is lower than film (at least it was in my E6 days).
Thanks for commenting - MC