I think the whole RNA thing is a bit over hyped. People like RNA because it looks somewhat like computer code, and our technology is just smitten with computers, these recent decades. And right along with that, it makes for such a painfully obvious way to implement inheritance and natural selection, a way to make genes.

But I think our focus is too narrow, we should think bigger than just obvious solutions. Now, it's great if the early oceans were awash with RNA, but does RNA exploit energy gradients on its own? The idea of all this random RNA just sloshing around, using nothing more than entropy, behaving as genes and inventing all the rest of it, just seems a little cart-before-the-horse.

What about metabolism-first? Is fire alive? It has many properties of life but, critically, lacks self regulation and heritability. Burns everything and goes out. Imagine though, a system of enzymes that can assimilate various specific kinds of molecules, and sustain production of those enzymes. Imagine this contained, compartmentalized in some way. Dynamically not much more complicated than simple fire, but crucially, the enzymes can be said to be inherited. These are not 'genes' as such, these are the actual machinery. It should be possible for natural selection to act on 'life' no more complicated than that. And that kind of system could then take find uses for RNA.

(And yes, the enzymes I refer to could be realized as RNA, but acting as the enzymes directly, not merely describing them, and this would confer some major benefits)