The placement you describe is emphasized in early instruction, because it often is necessary to leave adjacent strings clear, and because it is an exacting placement demanding assiduous care, which has to be internalized as an autonomous response through training, whereas a less than vertical approach in which there is or may be contact with (upper) adjacent strings is a more lax and intuitive approach.
But the this doesn't mean that one ultimately must rigorously adhere at all times to this application. The assiduous care it takes to maintain this approach can then be generalized into the assiduous care it takes to depart from the approach when it may be better to do so for a variety of reasons.
The following is a paragraph extracted from a much longer post I submitted to another forum many years ago, in response to a some steel-string fingerstyle guitarists who had been collectively targeting classical technique as somehow counterproductive to improvisational ability, which untenable thesis was to some extent attached to a view of classical as so burdensomely ritualized, that as a class, classical guitarists are unable to essay various technical devices in efficient fashion, or even with out actually injuring themselves.
The particular device addressed in the passage from which this paragraph is extracted was string bends, but as it pertained to fingertip approach, it can be cited here. The tone is somewhat disputatious, because that's the direction that the discussion thread had taken, but just so you know, it all ended amicably-- I think.
"....As for the finger position itself, the suggestion that a classical guitarist would not know how to position the tip of the finger for a bend is utter nonsense, and is tantamount to an imputation of idiocy to the entire class. "R---" may perhaps have been exposed to or been told of a pedagogical preference of some particular teacher to instruct one with absolutely no experience during the very first lessons to take care that the finger tip be held vertically in pressing down on a string, but it is sheer inanity to impute to the entirety of classical practice that this as an idealized "perfection" which is then artificially all-pervasive to everything a classical guitarist tries to do. As is practical in no matter what style one plays, the fingertip naturally can, and must assume a variety of attitudes, varying in accordance to the task at hand; with whether muting of adjacent strings is to be averted, is benign, or is deliberately employed; in furtherance of any lateral or longitudinal reaches; in preparation for a downward slur taking into account the extent of outward reach and the combinant effect of other simultaneous actions; lateral vibrato; and yes, in contemporary literature, bends. I know this because I do this. I use the same technique when playing steel as I do on nylon, and have never incurred tendonitis, even when playing straight through every day and into the evening at three and four day outdoor festivals. I know this because of colleagues and others who do the same. The monstrous mis-match that has been posited between classical technique and this particular device or the genre at large with which it is associated is entirely a myth. ..."