The meaning you appear to impute to my posts is so bafflingly different from the meaning I intend in them that I do not believe further conversation between us will serve any purpose.
Also, if you don't like rules interpretations that are a little out there you should probably avoid the "You Break it You Buy it" section, that kind of interpretation is the whole purpose of theoretical optimization (as compared to practical optimization).
I would disagree. There are TO builds without questionable interpretations of rules (e.g. much of Pun-Pun's cheese).
I'm more skittish than usual in this particular case because the rules actually tell you to be careful when improving monsters, and that the guidelines for CR (as opposed to the actual effect of advancement or templates on the monster's abilities) are only that. In particular, monsters with class levels and monsters advanced by HD beyond twice their original CR have a note that you "should modify its CR as seems logical"; I feel that this note, unhelpful as it may be, is still a rule and that holding up monsters in those regions as "badly-CRed" amounts to wilful ignorance of that rule (effectively, the official CR of monsters in those regions is "figure it out yourself", which is lazy but tautologically accurate). On the other hand, I see printed monsters, printed templates and suggested-in-monster-entry HD advancement to mostly be fair game here, as that's stuff they are
telling us to trust.
Speaking of which: advancing Spell Weavers by Hit Dice goes crazy very quickly, as their racial casting scales as HD+2. Dragon 338 actually published an Advanced Spell Weaver of Legend Archmage 3, which was allegedly CR 21 but cast as a 35th-level sorcerer with Improved Metamagic, two instances of Multispell and Epic Spellcasting. If you just straight advanced it to max it'd supposedly be CR 16 (or 17, depending on how you round it) but would still have four epic feats and 32nd-level casting.