I went to your pointer, then I went to here:
http://www.basketball-reference.com/about/ws.html
Both offensive win shares (doesn't even use Dean Oliver's formulas pre 1973) and defensive win shares are incredibly weak. Russell gets credit on D for minutes played and all the work he and his teammates did on defense. It's a stew, very little can be gleaned for sure in terms of attributes to which player and what value.
The offensive formula is joke because its aimed at this time, when high pace and average or slightly below average percentages are a bad thing. In Russell's time, missed shots were more common and less harmful because the other team couldn't shoot like todays teams either. So a team that averages the most or 2nd most points scored across 9 years gets dinged because it was 1.1 made FT away from an average shooting percentage? That's asinine.
Hahahahahah, I just looked at 1959-60 and OWS is a joke as all open minded souls will see.
Knicks: #1 NBA team in FG% at .421, Celts #3 a mere 4 one thousandths behind at .417. (lg/ave .410)
Knicks #2 FT team at .765, Celts at .734 in 4th place.
Celts #1 in PPG at 124.5, Knicks in 4th place at 117.4 (7.1 behind)
Knicks got 25.0 OWS and Celts had 19.1 OWS.
Multiply Celts FGA by Knicks FG% - 33 FG's difference (.44 FGM per game for the Celts and its a tie) - at 119.6 FGA per game, that's a pretty shrimpy number in the grand scheme.
FT - Celts need 78 to tie the Knicks (33.6 FTA per game by Celts)
So if we multiply FG's *2 so we can compare to FT's - we end up with 273 FT per game, and the Celts are 144 points on the season short (or 1.92 made points per game - less than a 2 point FG - one less failed putback by stone hands Russell per game (he had a ton every year) and our OWS will be above the Knicks) makes them equal - and this difference costs them 25% OWS on the season?
The 1959-60 Hawks are between the Celts and the Knicks, and have 25.7 OWS - more than the Knicks, with a lower FG% and lower FT%.
Focusing on efficiency in an era when few teams had it, and it was a legitimate strategy to run ones offense at a high pace/volume of FGA is like a historian judging the mores of Ghengis Khan by todays standards. It's misleading and it sheds no light or truth on the subject.
OWS in that era is almost useless and certainly sheds no real insight on the topic. DWS is the same thing because we are missing: blocks, offensive/defensive rebound split, steals, TOV's. Assists were handed out much more conservatively back than compared to today as well. Not enough info for % stats like USG, AST, REB, STL, TOV, BLK, DRtg, ORtg. It's even weaker than using BA, HR, and RBI for a hitter.