RNA and protein structures enable reconstructing ancient evolution because: 1. structures evolve more slowly than primary sequences; 2. Ancient self-organized patterns reappear spontaneously and/or 3. re-evolve secondarily. Previous analyses grouped RNA secondary structures in (a) presumably primitive, short RNAs rich in external loops (topping stems) and few bulges (unpaired nucleotides within stems); and (b) longer, more derived RNAs with more bulges, presumed regulatory endonuclease targets. These represent the main axis of RNA evolution from (a) tRNA-like to (b) rRNA-like. We suggest that relative similarity of tRNAs to (a) reflects antiquity, and to (b) the opposite, predicting that tRNA scores on this tRNA-rRNA axis converge with genetic code inclusion orders of tRNA cognate amino acids. This occurs in particular according to amino acids ranked inversely to tRNA isoacceptor diversity, and mainly in evolutionarily ancient organisms. Putatively, in some organisms, tRNA cloverleaves sometimes recover convergence with genetic code inclusion orders of cognate amino acids, probably because original ancient evolutionary processes integrated functional constraints, i.e., tRNA distinguishability to avoid misacylations. Results confirm the direction, evolutionary and biological relevance of the tRNA-rRNA secondary structure axis as RNA's major evolutionary axis, a potential calibrator of biomolecular evolution.