There is nothing that I would characterize as corresponding to FIIa/FXa for Lovenox, but Copaxone does do things that can be measured. Peptimmune and MNTA (it says) have measured them.
If giving a person a specified dose of branded Copaxone causes assay X to change from a baseline reading in the range of [B1, B2] to a reading in the range of [A1, A2], and giving a person a specified dose of generic Copaxone causes assay X to change from a baseline reading in the range of [B1, B2] to a reading in the range of [A1, A2], then the FDA will be more confident in concluding that the generic is sufficiently similar to the brand to approve the ANDA. It is not necessary for the applicant to understand the MoA of the drug to accomplish this (although such understanding would obviously help).
“The efficient-market hypothesis may be the foremost piece of B.S. ever promulgated in any area of human knowledge!”