Did you read my #123? How about this from that post:
I've seen some hypotheses offered by various camps: "all replicative biological systems will involve proofreading", or "bacteria will perform terraforming function for future evolution", or other such things.
The hypotheses may have to await technology for testability, but they are both clearly testable.
1) Knock-out proofreading systems from cellular replicative, translation, and transcription machinery. The ID hypothesis would be that evolvability will be lost, i.e. this altered population will shortly yield its existence to error catastrophe.
2) Identify the biological features important for the terraforming function. The ID hypothesis would be that these features are more important as role in future evolution than they are in whatever role they serve for the organism. Julie Thomas has an essay expounding upon this very hypothesis. It’s currently off-line, but I’m trying to locate it for you.
This seems to be equivalent to untestable. There's no indication that the technology will become available (nor indication that it will not.)
How does this distinguish ID from other other evolutionary theories?
What features are these? What is terraforming? The ability to kill everyone with bombs or poison gas or biological organisms? Beavers change the landscape (and atmosphere) as much as most creatures.
If there are no such features, does this refute ID? If these features exist but are not important (and how is importance measured?), would this be a refutation? There doesn't seem to be much testable here.