When the 1990 awards results were announced, Boston coach Mike Milbury said it was a joke that Bourque had been left off six ballots, and demanded to know which writers had left him off. Good on Milbury for supporting his player, but I don't think we need to take his point of view. I don't agree that a 1990 Hart ballot without Bourque - or a Hart ballot without Messier, for that matter - is proof of voting in bad faith, especially on a ballot with only three names.
RE: Leetch over Bourque in 1991-92, I think the statistical gap in favour of Leetch was too much for Bourque to overcome. Leetch scored 21 more points than Bourque, and the Rangers finished with 21 more points in the standings than Boston. Points and team success are important for the writers, and the gaps were huge. Bourque may have been better defensively than Leetch, but Leetch was much improved defensively under Roger Neilson and was no longer a one-way player.
Much of these gaps were from the last 2 months of the season. As of January 31, the Rangers were leading the league, but were only 2 points ahead of second place Montreal and 10 points ahead of 6th place Boston. And Leetch was 1 point behind Phil Housley in defenceman scoring and 6 points ahead of Bourque. At this point the Norris voting might well have been close. But from February 1 to the end of the regular season, Leetch outscored Bourque by 15 points, and the Rangers out-pointed Boston by 11 points and won the league by 7 points.
It may also be relevant that the league shut down for 10 days in the beginning of April due to a players' strike, and voters' opinions may well have been set at this time. At the time of the strike, Bourque's stat line was 77 GP, 19 G, 57 A, 76 P, and +9. A very good season by anyone else's standards, but Bourque was in danger of finishing the season with less than a point per game for the first time in a decade, and with a single-digit plus-minus for the first time in his career.