You're preaching to the choir on this to me for the most part. They are certainly overemphasized as measures of luck. I believe theres at least a significant portion of the pitching community who has control to some measure of both of these indicators. To be fair, even those who originally hypothesized the .300 BABIP/out of pitcher control idea backed off of it fairly quickly.I know it's against the recent "rage", but I think strand rate, hit rate, and babip are over-emphasized as measures of "luck". Each of these are, at least to some extent, skills. The ability to strand runners is an important element in a pitcher's success. Some part of that ability is luck, no doubt, but much of it is skill, IMO. If a pitcher can routinely get a double play ball or a strikeout when he needs it with runners on, he's going to have a good strand rate, and that's going to translate into good ERA and W's. I haven't seen an empirical study, but my sense is that there are some pitchers who deal with these kinds of situations better than others. The other point - it may be an acquired skill. That is, as a pitcher matures and learns more about how to pitch to situations, might this be expected to improve? Seems likely to me, although it may be hard to measure, as other skills usually fade with time.
The first challenge in empirical studies is what sample size to even use - most times you see BABIP of pitchers with 100, 150IP in consecutive years because of the issue of small sample size affecting the BABIP of pitchers with low IP totals. This is fine, except it basically limits the sample to pitchers who've already proven they can get major league hitters out to the point that a team would let the pitcher throw that many innings in consecutive years. So trying to apply those results to the entire pitching population has pitfalls already built in. I think there are plenty of examples of pitchers just coming up from the minors who don't have MLB quality stuff who get shelled for high BABIP. So I think inherently there is an issue even measuring the data correctly. To me, you have to look at it pitcher by pitcher and see whether there is a history of low BABIP like say a Mo Rivera (14 of 15 MLB seasons below a .300 BABIP) or a high BABIP (Andy Sonnanstine, 3 for 3 above) and apply that knowledge. That said, for an overall population, you regress BABIP to the mean .300 and for the whole you'll do much better than trying to estimate individual BABIP. I've tested it a bunch of ways and it always comes out that way. Doesn't mean you're doing justice to an individual projection that way, but to the whole, you're as accurate as you can be that way.
Strand is a little bit different because there's a clear tie in to strand against the 3 main peripherals (K, BB, HR) and BABIP. Striking guys out improves strand because obviously those runners don't advance bases on the batted balls that go for hits. Lower BB don't allow for extra baserunners. Lower HR mean the guys on base who automatically score on those HR. Lower BABIP means less hits to score the guys who do reach base. So untying strand from the otherwise strong peripherals is a real challenge. The original post that this comment was talking about Wandy Rodriguez' high LOB rate this season - and at 83%, it's probably not sustainable - his peripherals suggest he's closer to a 73%-75% rate and to me he's a high bet to regress.
I still think there's a window between about 65% and 75% strand and probably .275-.325 BABIP that is sustainable for the population of MLB pitchers and past that probably a projection system should reflect the normal range. (note that a guy like Rivera has a career BABIP of .266, Chris Young in SD is .256, Andy Pettitte .314) - the reality is most pitchers with high BABIP consistently won't stay in the majors so we don't have a great track record of high BABIP guys year in and year out, but there are a few. The real question is how much evidence we need from a pitcher's track record before we stop regressing them to the mean.
Thoughts appreciated, desired.