> Andrew: Is this the origin of all (or most) u-stem adjectives and
nouns? Does the presence of *-u- in adjectives always or usually
imply that the adjective at one stage had the suffix *-ro-?
It's a matter of internal reconstruction, and thus inherently disputable,
but it's at least one way of accounting for the *-u-/*-ro- variation
in deverbal (or deverbal-looking) adjectives. Of course other
scenarios may be closer to the truth in individual cases.
> And are such adjectives (and nouns) sufficiently distinguished
semantically from original *-o- adjectives (and nouns), assuming these
remained
No, the result would be the same, but we don't normally get _verbal_
*-o- adjectives with the zero grade of the root (while the o-infixed
type of *bHor-รณ- is extremely common).
> (i.e. is it necessary to have a contrast between *-us and *-os?)?
In PIE? Of course!
Piotr