Narby has stated flat-out that it's impossible, more than once on this thread. I don't claim that this is the "official" position of those opposed to ID, but it is certainly one that is loudly stated by some.
The only question of interest is whether it is necessary, and ID advocates have no curiosity about that.
That's an ad hominem statement, and probably also a strawman. But still, let's look at that "necessary" clause in light of the example we've been considering. Is ID a "necessary" condition in the case of insulin-producing bacteria? No, since one can describe a process by which it could happen without intelligent agents. Is "naturalistic processes" a necessary condition for the same example? Obviously no, because the right answer is that humans did it.
And thus your "necessary" condition fails in this specific example. One can likewise draw a more general conclusion that neither "naturalistic" nor "ID" hypotheses are "necessary." I'll go one further to say that "sufficient" also falls to the side once you enter a world where both naturalistic and "ID" hypotheses can plausibly explain observed phenomena.
But perhaps the word you're really looking for is parsimony: economy in the use of means to an end; especially : economy of explanation in conformity with Occam's razor. It
If ID is a serious intellectual proposition, it should be trying to demonstrate its own irrelevance.
A silly statement.
Parsimony is generally a good thing.
If anyone actually believes that irreducibly complex entities exist, they should be leading the research to reduce them. That's the way science behaves.