As I understand it, it's answering the question of why there's so much fine-tuning in the universe.
It's an explanation, it fits the observed data. It can't be tested, and the predictions it makes can't be verified. So until we can verify that new universes are created from black holes, with properties inherited from the parent universe, then it's just speculation. But interesting speculation.
And it's a real question that does need some kind of answer.
> It can't be tested, and the predictions it makes can't be verified.
To be clear, both Smolin's CNS and Gough's uptake of the theory make predictions and offer ways to be falsified, with Gough making accurate predictions of early galactic structure that Webb would see.
It's good to see people expecting predictions and falsifiability, but I'm curious why that requirement isn't being upheld against the standard model (which has been predicting more and more incorrectly) and string theory (which is unable to offer any predictions and also expects you to believe there are 9 or 11 dimensions or something). In my view, CNS requires the fewest logical leaps of faith.