Tightening up the bolt in the receiver of a Remington, whether by bushing, bumps, or a new bolt, appears to be good for about .040" of vertical at 100 yds. That's 1/25 MOA. The same seems to be true for Winchesters. When bolt clearance is really excessive (more than 6 or 7 thou) the improvement can be more noticeable since the detrimental effect of the sloopy fit is more obvious. I think it was John Fry, a Washington gunsmith, who had the first bushed bolt I saw.
On a Savage, the bolt and trigger design eliminates the deflection of the bolt by the contact between the sear and cocking piece so there is, possibly, less need for a decent fit.
Occasionally, one will see an action with an obvious alignment problem and the correction of the problem will have a marked effect on performance. Just as often, a crooked action shoots fine. After fixing, it still shoots fine. Likewise, we have all had perfectly good rifles which seemed perfect in all respects except that they didn't shoot.
I have always been going to set up an action so that I could deliberately introduce errors in the threads, bolt face, locking lugs, etc. and try to quantify the effects of various defects. This would be a very involved undertaking and I don't know that I'll ever get the time; or the money!
In many cases, action truing is much more effective in the shooter's mind than at the target. Regards, Bill.