Which is why that wasn't the argument. Both French and German cars have evidence of design (non-local improvements over earlier transportation), and both clear and garbled speech through a telephone have evidence of an intelligent agent behind them. In no case did I suggest making the inference that flaw -> not design. Rather, it's "being constrained to one domain of attraction in designspace" that indicates lack of intelligence behind the optimization.

>You don't refute ID by showing some sub-optimal function in an animal body plan, firstly because ID doesn't claim everything is optimal. ID implies intentionality, but does not and can not explicit intentions.

Specific variants of ID claim certain intentions, and the lack of intelligent optimization would refute intelligence being behind that intention. (To the extent that this intention is deliberately vague and kept from having observable implications, it can be rejected as non-scientific on that basis, without having to assert the claim, which I was criticizing, that the whole enterprise is unscientific.)

>Anyone can make an object that by all means can not be the product of a natural phenomena, and still someone else being able to tell it's design, could not really tell the purpose of it.

That just means inferences here (as everywhere else) are imperfect, not that detection of intelligence is impossible or unscientific. You can't know to 100% certainty whether the sound on the telephone is someone speaking to you or just a random coincidence of static, but that doesn't mean you throw up your hands, remain agnostic, and lambaste the ignoramuses who dare to claim they have heuristics for speech recognition.