Because science is about finding explanations, not asserting that explanations cannot be found.
How about ID coming up with a falsifiable test that doesn't rely on gaps in another theory? One that isn't really a test of the ToE and claim that if evolution can't explain it then ID wins by default?
Please describe this alleged "test for ID".
Only because ID has yet to actually observe anything that is "irreducibly complex". As Behe's examples of eyes and flagellas fall apart, he's resorted to redefining "irreducibly complexity" to be meaningless.
For example, if parts of a flagella have an excretory function, Behe claims that it's still "irreducibly complex" because excretory systems aren't flagella. But there's nothing in biology that says that proteins can't have more than one function or that proteins can't be put together to form a new function. It's very common.
ID'ers have yet to observe anything to support their theory and they really picked the wrong time in history to try this approach. Now that companies are sequencing and resequencing genomes, ID'ers examples will probably be shot down almost as fast as their thrown out.
Take a look at a company called 454 Life Sciences Corp. They're leasing a technology to Agencourt Bioscience that could result in human children being identified by their DNA at birth sometime in the next 10 years or so.
Because:
1. All specific examples of proposed "irredicubly complex" systems in living things so far have been found to not actually fit the specifications of "IC". Every time the IDers say, "this system is IC", it turns out that a) they have only *presumed* it's IC and not actually tested it to determine that it is, or b) the system is demonstrably non-IC (because simpler versions of it exist in nature and still function).
2. Behe's argument about how "IC" systems are not evolvable is fatally flawed. So even if something could be found that was actually "IC" as Behe defines it, it *still* wouldn't demonstrate that the system was "difficult to describe coming into being by evolution".
3. IDers frequently confuse "has not been explained by evolution" (usually because no one has researched it much yet, not because it's actually a conundrum) with "hard for evolution to explain". The two are not synonymous.
4. Even if something could be found that was specifically hard to explain via evolution, that does *not* constitute evidence *for* ID. It's simply a problem for evolution, which could be potentially resolved in any of a number of ways. In cases like that, science (and basic logic) requires a "don't know, needs more work" position on the topic, not a "ID (or any other alternative hypothesis) must be right just because evolution is faced with a puzzle" position.
5. Even ironclad evidence *against* evolution (if they ever actually find any) is not evidence *for* ID. ID does not automatically "win" even if evolution is falsified, no more than evolution would automatically "win" if ID was found to be flawed. After all, they could *both* be on the wrong track. Read this previous post of mine for more explanation of why that's the case.
The IDers have to find evidence *for* ID specifically in order to bolster their case, not evidence *against* evolution.