User blog comment:P進大好きbot/Evaluation of Analysis/@comment-30754445-20181120105029

It is important to note that a high Level does not guarantee that the analysis is trustworthy, and a low level (unless it's level 1 or a low level 2) doesn't mean that the analysis is useless.

Level 5 analyses are only as valid as the proof therein.

And level 3/4 analyses can sometimes be quite trustworthy, depending on the exact line of reasoning used and on the claimed strength of the notation (the higher the ordinal required for the analysis, the more careful one must be).

Also, I'm not sure why you put our classic analysis (Deedlit and co) of PS on level 2. We were using the extended Madore Psi, which was the bona-fide standard notation that was used in this community at the time. Just because the functions were not defined in our analyses themselves, does not mean that it is ill-defined or that there was any ambiguity regarding which version of psi we've used.

So those analyses are clearly Level 3 (not level 4, though, because no effort was made to ensure surjectivity).