[MadMax]

I have yet to see a comparison of groupings between a well tuned no hop up GBB and a TK twist barrel. I also find it wierd that nobody offers close tolerance slide and outer barrel kits. If you really wanted accuracy, you'd make your outer barrel close in diameter to the slide to reduce OB-slide play. Then you'd also closely match the diameter of the inner barrel to the OB to reduce IB-OB play.

All in all, it'd make sense to spring both the OB upwards (or downwards) against the slide for consistant registration, to the slide, and spring the IB against the inside of the OB.

It's also kind of bad that the slide has some play against the frame. The rear end of the IB is firmly secured into the lower frame which means that about half the variance in the impact point is a result of slide-frame play (lift the tip of your slide and you lift the tip of your IB). The sights are mounted to the slide so slide-frame play results in a variance in impact point relative to the slide (hence the sights). It's very difficult to make a zero play slide action (barring a goofy Sherrif bearing arrangement), so it make sense to install a spring system to bias the slide high (or low) so it at least registers the same way consistently so the rear of the IB relative to the slide ends up in the same place.

I'm not sure I have all that much faith in HK gunsmiths. They should be publishing grouping comparisons between stock build vs. their own changes if they were really in the biz of performance IPSC. Instead all I find is nice pics of custom cosmetic milling. Lots of projectile voodoo like TN or twist barrels, but when I look at my Hicapa with an aftermarket "Kimber" slide, I note about 1mm of play between the OB and the slide. The IB is closely matched to the ID of the OB at least (stock TM). 1mm play at barrel tip equates to about .4" of grouping variance at 33' (~10m). The slide adds about a further 1mm of angular play (lift at the front) which maybe doubles the barrel related variance to 0.8" at range of impact.

Tightening a group at 33' by 0.8" is a bit of a big deal really. I used to shoot target pistol a bit and could consistently break a score of 85 without too much practice (not really a big deal). A 0.8" reduction in group would push me past 90 consistently. 0.8" is the difference between a lot of outside of the "X" ring hits I I'm guessing IPSC score is similarly strigent.

Before chasing down the ephemeral effect of axial spin stability, it makes more sense to acheive consistent lockup, between boresight point and sight point, when the slide goes into battery.