I agree that example 5 on MSU is not an example of kinetic v/s thermodynamic control, because what acetic anhydride basically does is - it attacks at the less hindered site, to produce the less stable initial dimethyl carbocation, which rearranges and stuff.
However, look at the first example. Or the one that i posted in my last reply. Here, the website, and some books claim that thermodynamic v/s kinetic product control exists. The examples listed on MSU are taken directly from the Solomons and Fryhle's O-Chem book. I posted the MSU link since it was inconvenient for me at that moment to scan and upload the pages. Now, the Solomons text book has the exact same examples- but the mechanism is not worked out - while MSU has given a mechanism, but I have serious doubts about its correctness.
An important thing to note is , Kinetic v/s thermodynamic control has been shown to occur only when there is hydrogen as a substituent on the beta carbon. None of the other cases, including the one I originally posted in my first post have been reported/stated anywhere to show KFP V/s TFP.
Now, this intrigued me big time. From your discussion, and also after reading many articles and chapters on Thermo v/s Kinetic control, I got it clear that for an alternative TFP to exist , there must be a plausible reaction mechanism from the TFP to the KFP, otherwise how will the initially-formed-faster-major-product get converted to the TFP?
Standing here, what I feel is , the mechanism shown on MSU is not correct. It contradicts my understanding of chemistry. I mean, why will a protonated carbonyl oxygen lead to a carbocation? I mean even if one applies Resonance, the resonance contribution by the carbocation will be way too small in contrast to the octet-stabilized protonated carbonyl. So a reverse reaction even if it occurs, will be exceptionally slow if that is really the mechanism. And I don't so why such a reaction should at all take place!
I think a possible and more logical explanation can be given with keto-enol tautomerism . In example 1, initially an aldehyde is formed. But it is an rapid equilibrium with its enol form. Now, normally the -OH of an enol form does not get protonated because it would lead to a highly unstable vinylic carbocation. However, in this case I will argue that protonation would still occur, but that would not lead to a carbocation, but a resonance stabilized Phenonium ion - Neighboring Group Participation by the Phenyl ring. Now, when we visualize the phenonium ion transition state, the bonding is unsymmetrical , and more amount of partial positive charge is present on the carbon atom which is attached to two phenyl rings.
The molecule(s) of water lost in course of the reaction can attack using their lone-pairs on oxygen at the delta-plus end, resulting in the expected phenyl ketone.
This is my explanation/ how I would like to justify the reaction. Once again, I think the mechanism on MSU for Example 1, is highly improbable.