The way I see it, which is based on my experience, a TRT in a fussy reader is the most sensitive test of a disc's actual reading condition. Then comes the Scandisc, and finally the PIE/PIF scan, which often tells nothing at all about the actual reading condition of a disc, as it only reports some numbers about low-level reading errors. Useful for comparing burning quality etc.., but useless for checking actual performance.
Of course, drives that report POFs accurately during a PIE/PIF scan, turn a PIE/PIF scan into a far more useful test for real-world behaviour! Sadly, many don't.

Also, the fact that so many users scan at slower speeds (@4X, @8X), while being kinda backed up by scanning theory and ECMA standards, makes most PIE/PIF tests absolutely meaningless in terms of actual reading behaviour, which is nowadays expected to be flawless @12X, 16X, and even @18X in some drives. It's easy to pass a @4X or @8X scan, (= show in-specs numbers), it's another story to pass a @16X TRT in a picky drive without any slowdown.
So the best discs/burns (in terms of
current actual reading performance), are not those that show the lowest PIE/PIF figures in some end-user home scanner, but those that show a perfect reading curve in a Lite-On 16P1S or 16P9S DVD-ROM, or a NEC 4550A burner, or some other fussy reader.
On the other hand stability, consistency, low-level burning quality, are better checked with one of the usual consistent PIE/PIF scanners. TRT's usefuleness, in this domain, is extremely limited, not to say close to none.