I do think that the new compound node will be great for complex materials, where much of the inner workings (the node spaghetti) "simply" encodes known light physics equations, in Poser node format... and well, basically isn't needing to be exposed to the end-user...

...this is much of the point of the compound node, isn't it?

So, would make sense to be incorporated into Matmatic, wouldn't it?

Looking at BBGlossy, for example, I guess all but the left-hand side parameters, could be tidied into a compound node, quite happily, couldn't they?

Although, I have to admit, I do rather like gazing in wonder, like some primitive proto-human, at the bewilderingly awesome tangle of nodes to the right of the main node there... ;)