It certainly makes a lot more sense than trying to calculate the odds of some specific protein fold emerging de novo.
Evolution doesn’t have to evolve some particular protein fold on command. In order to explain how a cell can come to have a genome with thousands of protein coding genes there just needs to be some way for novel protein coding genes to evolve.
That means there has to be enough selectable function in protein sequence space, for random expression of genetically encoded open reading frames to occasionally result in a functional, adaptive, polypeptide chain. This has been experimentally demonstrated to be the case.
It is of course also strongly implied by comparative genetics, with which it has been shown that there have been a relatively steady influx of gene gain and loss over the course of the history of life.
That means the two different “schools” of evidence corroborate and support each other. Direct experiment and historical inferences both show the same thing: Protein coding genes of all types can and do evolve. Whether they serve as macromolecular structural scaffolds, bind and stabilize cofactors, or each other, whether they are fat or water soluble, or catalysts of chemical reactions. They are evolvable entities.
Novel functional protein sequences are found in random explorations of amino acid sequence space, they have adaptive functions, they are retained by and their degree of function enhanced by natural selection, they become shuffled and assembled by recombination and fusion, grow into larger and more complex folding domains, and ultimately can assemble into elaborate multi-component molecular machines.
But one problem here is the misguided and simplistic thinking of ID creationists concerning protein evolution. You have this strangely constructed picture where a particular gene is a sort of target that must be experimentally recreated in de novo evolution before your very eyes otherwise you won’t believe it. It’s weird because you hold this standard in no other area of scientific historical inference. You’re not demanding that geologists re-grow particular (or even any) mountains before you accept that the forces operating in the Earth’s crust and mantle is the sort of thing that can produce the mountain ranges we see. On those subject, you are entirely fine with inference to the best explanation for highly specific and unlikely outcomes.