Are gaps important in design inferences? Most ID proponents seem to think they are. Usually an ID argument tries to show two things:
(1) Natural (non-intelligent) processes (probably) cannot account for x.
(2) x closely resembles designed objects.
(3) Therefore, x is probably designed.
But what if (1) were false? What if
(4) Natural processes not only can but do account for x?
Does that mean that x was not designed?
No. For either
(5) The designer could have either introduced a previous designed object y, knowing that natural processes (probably) would then take y and produce x.
(6) The designer could have produced the natural processes themselves, knowing that they (probably) would in turn produce x.
And if (2) is still true, then the design inference should work.
So really, showing gaps is only one of the ways to construct a design inference.
Mike Gene seems to rely upon (5) in constructing his design inferences.
Thomists seem to rely upon (6) in conducting their design inferences.
Of course, I don't think there's an inherent reason why proving (1) is a bad way to go.