I'm glad I saw this Q and A, I wasn't aware of the redirects issue. After reading the explanation, my first thought as a computer scientist is why aren't redirects updated rather than chained after a merge?
For example, merging Schmoe-3 into Schmoe-2 redirects 3 to 2 as usual, but when Schmoe-2 gets merged into Schmoe-1, 2 gets redirected to 1 again as usual, but instead of leaving 3 redirecting to 2, update 3 to redirect to 1.
Of course, I don't know how the internals of wikitree work so there might be a logical reason why my suggestion wouldn't work. Perhaps searching for all redirects to a given profile is expensive? If so, maybe instead of updating redirects at the time of merge, a batch process running periodically could seek out redirect chains and fix them later?