Frankly, the information that is necessary for the examiner to recognize a best mode infraction is rarely available to the examiner.
The examiner cannot examine your inventor notebooks, your internal documents, your e-mails, etc.
Example - a best formulation requires lithium bromide and it works better than lithium chloride or lithium fluoride, but the written description only discloses the broader category of the lithium halides - so best mode requirement has not been satisfied. The only requirement is that the best way is disclosed.
The best the examiner can do is by careful reading of your disclosure whether you knew that one mode was superior to another, and if so whether you then disclosed that mode well enough to enable one of ordinary skill in the art to practice that preferred mode.
In my years of practice I have never had an examiner challenge me on this requirement.
The EPO’s new matter is essentially a super strict version of written description where you basically need the precise words and also the precise combination of elements recited. Your “If the original filing anticipates the claim” test looks like the “novelty” test the EPO uses, to check whether a prosecution amendment adds matter.
The public policy reason to stop Applicants getting to issue with claims that really ought not to enjoy the filing date of the app as the date of their defence against validity attacks is the same in both jurisdictions, yet Americans think EPO is super-strict while Europe thinks the USPTO is super lax.
Accordingly, primarily relying on the Examiner’s lack of fully developed reasons for the rejection will, IMO, often result in an affirmed rejection by the Board.