That could easily be solved by using league performance rating, rather than having a seasonal rating. The amount of wins or games is not important, rather the percentage of wins matters as well as the average rating of your opponent.
Don't you think that would mean more avoiding and activity decrease? Do we have enough players to separate high and low rated players?
As I've mentioned, that is a possible downside to using league performance rating. LPR is still something you can improve on by playing more games and winning, though. But that will depend on who you're playing. If you only play people with low rating, then your winning percentage will go up, but the average rating of your opponent goes down. (one factor increases LPR, the other decreases it). But the problem you want to solve is that people who aren't that active because others can play TUS far more actively have an edge with regards to playoffs. This will reduce that edge. People who play a lot of TUS now will still play a lot of TUS if we implement LPR, though.
A couple of things could be done to still keep people active if avoiding is to become a problem - we could hide the league performance rating until it's time for playoffs (gives another edge of tension, doesn't it?) or we could impose activity rules - no games in the final week (or 2 weeks) - no playoffs.
As for having enough players - TUS has been active for 6 years. By now, ratings should quite accurately display how strong someone is at certain schemes (except for people who only recently joined and whose rating has not yet caught up to their actual skill)