Just an FYI: The reliability of the Western Blot in HIV testing has had a similar history of problems.
Antibodies are specific to epitopes and NOT antigens. An epitope is a portion of a molecule to which an antibody binds. Each Borrelia surface antigen has multiple epitopes. Each surface antigen has a set of epitopes. Antibodies are produced by B cells with paratopes that match and bind to each epitope. So there are typically multiple antibodies with unique paratopes matching unique epitopes on each surface antigen. OspC seems to be the surface antigen with the least variation in epitopes which has made it a target for a "better" test since the epitope set variation is the basis for the Western Blot strain or species sensitivity. The B31 strain has a very specific set of epitopes on each of its surface antigens. A widely genetically diverse strain from B31 or another species will have a different epitope set for the same surface antigen.
Antibodies bind to epitopes on antigens and not just antigens. So a widely diverse strain or species with a very different set of epitopes will lead to Western Blot blot binding failure to the primary B31 based antibody. The B31 primary antibodies have paratopes that match the B31 antigen epitopes. If the surface antigen has sufficiently different epitopes, then fewer antibodies will bind during blotting. This lowers the band darkness reducing sensitivity. Only the antibodies spread out into bands by gel electrophoresis that have epitopes matching the B31 antigen epitopes will blot. This is the underlying reason that using the US B31 based Western Blot fails to detect European species. It doesn't just apply to species. The distinction between a Borrelia strain or species is not based on its surface antigen epitope variation. As the genetic diversity increases as in strains or species, the surface antigens epitope sets eventually vary enough to cause the Western Blot to fail. This is when the test antigens are replaced with ones that match the Borrelia in the area such that this epitope set diversity is reset.
There is no magic epitope set difference boundary between Borrelia burgdorferi, Borrelia afzelii , Borrelia garinii, Borrelia miyamotoi etc... As the genetic diversity increases, the epitope set differences increase. Eventually this causes the blotting failure when the blotting primary antibody paratopes fail to bind to valid antigens. This is not a linear function since the immune system can prefer certain epitopes making them more dominant. But eventually, the match breaks down, the quantity of antibodies that match during blotting decreases and the test fails.
This is the reason for example, the tests outside the US use antigens from local Borrelia. All the Borrelia share the same set of surface antigens ErpA, REVA, DbpA, OspC, OspD, 30, OspA, OspB, BmpA, FlaB, VlsE, 45, BBK32, 58, 66 and P83/100. Unfortunately, the antibodies for each of these can differ by strain or species. This is because the epitopes on each of these vary as diversity varies. Otherwise it would be easy to have a universal test.
It is this underlying paratope->epitope matching across a whole set of epitopes on each surface antigen that is where the problems lies. This is also why the C6 test was pursued since its epitopes are highly conserved across strains and species that gave it some advantages. Even the C6 fails since the epitopes on C6 also vary making it fail for some cases.
Since most of the studies exploring the CDC 2T and C6 were done in the US Northeast, the underlying Borrelia strain diversity was biased leading to a bias in test results. Unless the studies are done with known underlying strain diversity, this epitope set problem will lead to biased results. This is why nearly every country has its own 2T like test with local antigens with closely matching epitopes and why the criteria vary.