The ID argument appeals to comp sci people like myself because we see an analogy between DNA and computer code. Evolutionary algorithms are pretty bad at generating useful code, even with a whole lot of intelligent guidance, hence our distrust of evolutionary theory claiming to produce even more spectacular outcomes with little to no guidance.
The feedback I usually see is that DNA is not computer code, and it is a false analogy. However, either (1) the whole DNA -> organism process is Turing reducible (thus DNA is computer code), or (2) it is not.
If (1) is true, then well known theorems like the law of information non growth means unguided (and probably guided) evolution cannot generate highly complex organisms through mutating DNA. This is identical to the ID claim that chance + determinism cannot generate CSI.
On the other hand, if (2) is true, then there is some super Turing process at work. This turns out to be identical to the ID claim that intelligence is necessary to generate CSI.
So, when considering the computational nature of DNA we are faced with a dilemma, and either alternative proves ID correct.
Thus, according to the law of disjunction elimination, ID must be correct, regardless of whether DNA is computational or not.