So, for comparison, in your previous example (post #1, from bar 2 to bar 3) you had no actual parallel octaves, but an accented one on bar 3. And although that is a little bit mitigated by the second half of bar 2 being only a re-voicing of the first half, the accented octave is still a little strong since they are in the outer voices and they both leap.
You can find the odd accented octave or fifth in outer voices in Bach's chorales, but I don't know if there are any where both voices leap, though it is possible. IMV, that example is kind of borderline in a Baroque context. And in Baroque practice, continuo musicians sometimes used diminution to smooth out the leaps and/or to provide more activity to the base structure, in which case here the performance could produce actual parallel octaves. And the mind's ear, for some people, will perform it's own sort of diminution, which is why the octave or fifth in outer voices is overall heard as "accented".
But in the example you just posted, in a polyphonic texture of maximally independent voices, going from bar downbeat to bar down beat, the "afterbeat parallels" would be regarded as disruptive to the goal of independence, since that one slice of intervention on beat three is not enough to avoid that. With three beats of intervention, the parallel is largely considered to be non-disruptive.
In homophonic keyboard textures, of course, these types of parallels are usually not a concern since the texture is not one of maximally independent voices.
__________