For two points very close to being opposite each other on a sphere (the oblateness of which is so small as to be irrelevant) claiming there should be up to 12% difference between routes relatively close together is genuinely absurd.
The shortest (great circle) distance between the two locations in your example is ~19700km (evidence below). The other part of the same great circle is the longest distance. If we take the Earth’s circumference to be a nominal 40,000km, that would be 20,300km.
The difference between the genuine shortest and longest distances between these two points is therefore about 3%. How can you expect to see a difference of “7-12%” on routes so close together? It makes no sense.
You can’t just put people (everyone, apparently) down because of your claimed superior knowledge of maths. There seems to be something fundamentally wrong with your thinking on this subject.
Also, here’s a composite of your example from google maps. The relatively small differences are consistent with MSFS…
Google maps also lets you put the midpoint on the opposide side, creating the longest route. Absolute maximum distance possible appears to be ~20,300km which corroborates the numbers above…
