Well, you can probably guess that this is a loaded question...
The most accurate springer I have, bar none, is a break barrel (Pro Elite) - admittedly with a lot of tuing, but then I'm only comparing it with other highly tuned springers. The 80 I have is also excellent - on a par with the TXs.
I also noticed something interesting when rebarreling a TX.. the tension in the lock up ball bearing catch does have an affect on POI - around 1-2" at 30 yards.
So where I'm headed is that maybe non-tensioned lock ups (e.g. a sliding catch, not a ball bearing, like on the early 77) might be more accurate on underlevers ? It's hard to say how much variance there may be with the ball bearing, but if it gradually rotates, has gouges in it, isn't perfectly spherical, or the spring behind it moves a bit, maybe it could be pushing the barrel a little out as with my "manual introduced error" ? Might try magnetic retention - it's certainly cheap and easy
I'm also wondering if a full free floating barrel (i.e. with no underlever
loosely connected to it) - so either a sidelever, or a concealed underlever, or a break barrel - might be more consistent in terms of harmoics / vibrations ?
Sounds like a project for Jim.....