In the murasama ans 125 .tyre files there's no r_Bx3 parameter (hence it's zero), while in the varese's .tyre files r_Bx3 is explicitly set to zero.
When r_Bx3 is zero then the longitudinal force generated by the tyre does not depend on the camber angle (bike lean), even when there's a sideslip angle (and hence a lateral force generated by the tyre ). That's equation 10.E54 in Pacejka's book.
This seems strange to me, no ?