I think the cost of their investment potentially being decreased by leakage to be enough for the average masternode owner to abstain from data sharing, but I can't speak for every owner out there. Further, the usefulness of even fragments of data aren't nearly as helpful as the number of rounds is increased. Unless you have enough corroborating owners sharing what inputs they have per round, you're still at the mercy of probabilities as it relates to having a complete data set for the full beginning to end flow through. Unless you have the pieces that relate to inputs and outputs for the full set of rounds, you're still left with assumptions on whose address belongs to who. Could you draw some meaningful conclusions, perhaps, but hard to quantify without a real world scenario (not to mention blinding isn't even in effect).
As I said earlier, I believe in a mature scenario, essentially every masternode will be sharing data. The economics dictate it, because the data has value and masternodes that don't realize that value will be uncompetitive.
Sure, today some people run masternodes because they believe in the cause, and they won't spy, but once it becomes a business, all bets are off. (Also, let's be honest about it, today the data has no value because there is almost no real usage.) Money will flow, and so will data. Nobody really expected the degree of an internet data tracking industry (ignoring the whole NSA part of it) that exists today, but its not surprising at all given the incentives.