Also, it is a logical necessity that the transcription of RNA into proteins, including the ribosomes and the transfer RNAs appeared only later than the replication of RNA.
The part of the RNA world hypothesis that seems unlikely because it has very few supporting data is that the original biocatalysts a.k.a. enzymes were RNA molecules unlike anything that exists today.
In my opinion, it is far more likely that the first enzymes were non-ribosomal peptides, somewhat resembling to parts of the proteins that constitute the enzymes of recent life forms.
ATP and other nucleotides must have existed before RNA as a means to perform condensation (i.e. dehydration) reactions and the copolymerization of the nucleotides must have appeared as some undesirable side reaction, which somehow at some point generated some sort of self-replicating RNA, which was in fact the first virus, because its synthesis could not have any beneficial effects for the hosting life form.
Only after the appearance of the transcription of RNA to proteins, the RNA could evolve into a source of components for the hosting life form so that RNA could change from a parasitic entity to an essential part.
Life forms with RNA (and later with DNA) are analog with microprogrammable processors, because the nucleic acid molecules are memories which direct the synthesis of other components, so you can change the components by writing a different information in the memories, without changing the synthesis machinery. This kind of flexibility was certainly necessary for further evolution.
Life forms before RNA are analog to hard-coded processors. Supposing that they were made with non-ribosomal peptides, there must have been a mechanism by which a complete life form was replicated, by a cycle of reactions were some kinds of peptides were assembled by other kinds of peptides.
Non-ribosomal peptides exist even today in all living beings and they are assembled in the correct sequence from amino-acids without depending directly on the information stored in nucleic acids, like the proteins.
Memories, like nucleic acids, are certainly not necessary for designing an ensemble of molecules able of self-replication, exactly like a memory is not necessary for implementing the control automaton of a processor.
Nevertheless, it is likely that the further evolution of such a primordial life form was almost impossible, because all accidental changes in the structure of the component molecules would have resulted in a replication failure.
After the separation of functions between a memory (nucleic acid) and a generic synthesis machine, evolution by changing the memory content became much more likely to be successful.