I think we generally agree on this. I'm not dismissing baserunning and defense outright, I just think it should represent a smaller chunk of WAR until we can measure them as accurately as we do hitting. Thus your approach of halving the defensive values is one I can get on board with, and thus why you must agree with me that defense has too large of an impact on current WAR calculations.
Yeah I can get behind that, though I also think it's conceivable players like Machado were able to save 20+ runs in the field in a single year, so it's all up to individual interpretation in the end for defense. I think there is some merit in using a combination of UZR/DRS/Inside Edge.
I think we'd have to agree to disagree on the baserunning and positional adjustment components though. As great as a season as Martinez had, you could probably find a slow plodder in AAA who could manage a +100 wRC+ and be a replacement level DH. Not the same for a SS or CF.
And to take on your notion that SBs are team-specific, I see your point, but if Revere's being held at 1B on the Jays, it may show up in extra in UBR that he's then able to grab 3B or run to home from first. It'll also go the other way if a runner's allowed to run too much... It'll never be exact, but I think you're being picky about the edges here. Baserunning tends to be a fairly small component (and if a runner's fast enough, he'll damn well earn the freedom to run more).
Just by my quick-n-dirty method, halve Bourn's (and Martinez's) fielding runs and he's only 0.5 WAR better than Martinez that year