YOur essential point seems to be that we dn't knwo enough about the flagellum to directly identify whether it was designed or jsut evolved.
I agree.
But this is a MAJOR problem for the ID supporters.
You see, ID rests - or used to rest - on attempting to prove that certain systems were designed. But the problem is that we don't have any direct evidence for that - ID relies on trying to refute other hypotheses.
So, why pick on systems where we *don't* have a lot of direct evidence that lets us work out how they appeared ? ID's eliminative approach is - as your post recognises - at its weakest in that case. Remember that it was the ID supporters who chose to focus on "the" (or rather *a*) flagellum.
The nonteleological explanation has the advantage. It does not require assuming a designer. Without evidence for a designer it has to be preferred.
Which leads to another odd thing about ID. EVERY other design hypothesis I know of makes some assumptions about the nature and capabilities of the designer. SETI does, archaeologists do, forensic investigators do. ID doesn't. Why ? Again it weakens any argument for ID by limiting it to an eliminative approach, as Behe and Dembski do.
Mike Gene tries to do better but without much success and his hypotheses don't seem to offer anything very different from panselectionism. Both predict function almost everywhere. But he is still limited by avoiding hypotheses about potential designers.