Thanks, Z. I did take SF's to be more gut feeing due to her language ('it seems that...', e.g.). Your response countering her on relievers/innings was more direct (no qualifier), which suggested that you had some substance backing it. You also began your next thought (Joe's managing BP availabilities) with "it just seems like", further calling attention to that initial lack of qualifier. So it came across like you had something more than gut feeling.
On data - with all the analytics present in today's game, I think data on reliever performances is out there, it just may be laborious to mine the specifics to fit scenario. (I spent a few minutes on it this evening, but was stuck with U2 still having not found what I was looking for.) That said, relievers tend to be erratic. So your point on 'relevant' statistics still merits consideration given sample sizes, etc.
End of day, I agree we're arguing for the same thing - best reliever on the mound at a given time - just approaching it from slightly different positions. For me, the members of this bullpen have been maddeningly unpredictable from one appearance to the next, even by general reliever standards. So I think consideration should be given to sticking with one who happens to be 'on' so long as their overall pitch count is within range. But even if Girardi were to prove successful in knowing when to hold and when to fold with that strategy, I think it more a maneuver to better tread water in the short-term - until the black cloud that's been camped out above the Phillies' bullpen decides to finally move on (or the offense keeps performing like they did tonight).