i agree - kinda antithetical to the ethos of Pure Data, really. i think you could use Heavy to turn a patch into rather voluminous C++ code. i remember from the devs that they said the code output wasn't supposed to be edited or analyzed by humans, but rather existed to be wrapped to work in a number of environments as sort of a black box or code blob. but obfuscating a PD patch itself seems stupid. i'm sure you could name subpatches and abstractions randomly and maybe even rearrange the physical placement using the method Jonathan refers to, though i believe that could affect signal flow and execution order, wouldn't it?
best,