In this paper, we suggest a biomathematical numerical method analysing mRNA nucleotides sequences based on UA/CG Fibonacci numbers proportions. This method is used to evaluate then compare the spike genes related to the main SARS-CoV2 VARIANTS circulating presently within the world. The 9 main results proposed to be reproduced by peers are: 1/ SARS-CoV2 genome and spike evolution in one year 2020-2021. 2/ SARS-CoV2 Origins. 3/ Comparing 11 reference variants spikes. 4/ analysing 32 CAL.20C california variant patients spikes. 5/ Toward a meta mRNA Fibonacci gene end message code. 6/ analysing S501 UK, S484 South Afrika and « 2 mutations » IINDIA variants. 7/ Suggesting a possible variants spike mRNA palindrome symmetry metastructure improving mRNA stability then infectuosity. 8/ Analysing Fibonacci Metastructures in the mRNA coding for the vaccines PFIZER and MODERNA. 9/ Does the CG-rich modification of the synonymous codons of the spikes of the 2 mRNA vaccines affect the expression and quantity of SARS-CoV2 antibodies? Particularly, we suggest the following conjecture at mRNA folding level : CONJECTURE of SARS-CoV2 VARIANTS: The growth of long Fibonacci structures in the shape of "podiums" for almost all of the variants studied (UK, California, South Afrika, India, etc.) suggests the probable folding of the Spike mRNA in the form of a "hairpin", which can strengthen the cohesion and the lifespan of this mRNA. Finally, we show that this kind of Fibonacci matastructures disapears TOTALLY analysing the published mRNA sequences of PFIZER and MODERNA vaccines. One fact is certain, the 2 mRNAs of the Moderna and Pfizer vaccines will result in a low functionality of the spike vaccine because by doping these sequences in CG rich, their designers, in search of greater STABILITY of these RNAs will have built, according to us , sequences which, as soon as they are inserted into the human host, will seek to mutate, like SARS-CoV2 variants, towards CG ==> UA forms in order to improve, paradoxically, their STABILITY and probably also their LIFETIME.. Particularly, using new biomathematics theoretical methods (Master code and numerical standing waves), and comparing the Spikes of the 2 vaccines Moderna and Phizer, we conclude a very probable difference in stability and shelf life of the 2 respective mRNAs of these 2 vaccines. However, the “State of the Art” will tell you that their 2 protein sequences are strictly identical. However, by having modified their synonymous codons using different strategies, no one can guarantee that the quantity of antibodies generated will be identical in the 2 cases. We can only note the great ADAPTATION power - at the global scale of their genomes - of the most infectious VARIANTS such as the BRAZIL 20J / 501Y.V3 variant (P.1). This is very worrying for the VACCINES <==> VARIANTS run: We demonstrate how the Brazilian variant P.1 which becomes uncontrollable in Brazil in April 2021 has a level of organization of long metastructures of 17,711 bases covering the genome which is 3.6 more important than that of the 2 reference genomes SARS-CoV2 Wuhan and worldwide D614G. We suggest that this high level of overall structure of this variant contributes to the stability of this genome and, possibly, to its greater contagiousness.