Any software engineer can tell you that modularity is an essential characteristic of any program that must be actively developed, or even just maintained. Without modularity, any small change to one piece is liable to break a completely different piece in unpredictable ways. Adding a new feature to an unmodular system is difficult because it requires making many simultaneous changes to various pieces of a program that have no obvious relationship to each other or to the feature being added.

The same thing is true with Biology. In fact, it isn't controversial that the more complex an organism becomes, the less likely it is to evolve novel function, largely because of the intricate dependencies that must be maintained. What hasn't been attempted, so far as I know, is a quantification of approximately how much room an organism has in which to evolve – call this "mutation space" – as a function of its interdependecy complexity. In other words, how much functionality could be added/edited without requiring an unrealistic number of simultaneous compensatory changes elsewhere?

If it could be shown that any reasonably complex lower organism did not have room in its mutation space for the sort of evolution required to produce higher organisms (that is, any introduction of novel function would require an unrealistic amount of compensatory mutations to get off the ground), it would provide incredibly strong evidence for ID.

I'm not sure exactly how to quantify interdependency complexity and mutation space, but it seems like there ought to be a way to do it. Suggestions and/or reasons why this is a nutty idea that will never work are welcome.