In other news, the recently elucidated thermodynamics of self-replication shows that RNA is the only molecule known to be useful for the task.

RNA passes the self-replication heat bound while it is stable enough to form cells with lipids. DNA/TNA et cetera modifications, and proteins/PNA et cetera modifications are too stable. [ "Statistical Physics of Self-Replication", England, arxiv 1209.1179, to be published.]

And cyanobacteria is known to be very derived. About half the history of Earth passes before archaea and later bacteria appears as clades.

Protein fold family trees shows that, and that all proteins are derived out of RNA because the first fold couples to some of its nucleotides. [The latest rooted tree and the fold history would be in: "Giant viruses coexisted with the cellular ancestors and represent a distinct supergroup along with superkingdoms Archaea, Bacteria and Eukarya.]