I believe your quote is from the article posted here. I was talking about the paper, which I only read lightly.
Checking now the Results section of the paper, they do address how conservative the relevant genome site is, extracted below (emphases mine). TLDR: very but just short of completely conserved. Seems hopeful.
---
> Comparison of SP1-77 footprint to the mutations on the RBD of variants.
When the SP1-77 footprint was projected onto modeled RBDs from different SARS-CoV-2 variants (Fig. 5, A and B), it was evident that most mutations were located outside of the SP1-77 binding site, except for the Arg346Lys in the Mu and Omicron-BA.1.1 variants, and the Gly339Asp and Asn440Lys in all current Omicron sub-variants (50).
The modeled interfaces of SP1-77 bound to the Mu and Omicron RBDs, based on the SP1-77/G614 S complex, indicated that the conservative Arg346Lys mutation would preserve the interaction with HC CDR3 and would not have much impact on binding. Likewise, the Gly339Asp and Asn440Lys mutations were at the edge of the SP1-77 epitope with their side chains pointing away from the binding interface, explaining why Omicron variants were sensitive to SP1-77 neutralization (Fig. 5C).
While mutations near the Asn343-glycan in Omicron reconfigure the orientation such that the carbohydrate projects away from the protein surface (51), such mutations had minimal impact on binding and neutralization activity of SP1-77 (Fig. 5C).
Thus, these studies provided a structure-driven hypothesis for why SP1-77 potently and broadly neutralized all tested SARS-CoV-2 variants through Omicron BA.5.
Finally, we compared the SP1-77 footprint with the new mutations on the recently emerged BA.2.75 sub-variant. The N460K mutation in BA.2.75 was far from the SP1-77 footprint (Fig. 5, A and B); while the BA.2.75 G339H and G446S mutations were close to but did not overlap with SP1-77 footprint as described for BA.1.
These findings made us hopeful that SP1-77 also will bind and robustly neutralize BA.2.75.