Why are you obsessed with quantitative evidence? No one can predict what is going to happen to steem after such a change. It's all about trial and error. That's why SMTs are important.
We can try taking an educated guess to make good changes, but those guesses come from experience and not statistical models.
The blockchain at this point should move out of "testing" phase and into the real world. There's enough real world use-data that could be used that no "test" should ever be conducted so blindly. And I dissent your opinion that this would have no "known" negative affect on the blockchain.