I don't know, why anybody is insisting on this confusing idea of relativistic masses (in fact when you introduce a relativistic mass it were even direction dependent, which has been said already above or recently in another thread in this forum).
From a quantum-theoretical point of view the concept of mass is indeed pretty different in relativistic and non-relativistic physics. In non-relativistic physics of an elementary particle it is introduced in the construction of the quantum theory from the analysis of the underlying space-time symmetry (Galilei-Newton spacetime) as a central charge, extending the "classical Galilei group" to the "quantum Galilei" or "Wigner-Bargmann group". In special-relativistic physics the underlying space-time symmetry is the Poincare group (and only the proper orthochronous part connected smoothly with the group identity!), which has no non-trivial central extensions and central charges. The result is that mass (squared) is a Casimir operator labelling the possible unitary irreducible representations, of which those with ##m^2>0## and ##m^2=0## lead to a physically interpretable quantum dynamics and particularly the successful local relativistic QFTs.
The conclusion, also valid for the macroscopic world, is that in special relativity there's no additional conservation law for mass, and conclusively mass is not an additive conserved quantity. All conservation laws in special realivity related via Noether to the space-time symmetry are the 10 conserved quantities energy, momentum, angular momentum, center-of-energy velocity.
So from both a mathematical and a physical point of view one should define mass in special relativity as a scalar quantity and energy and momentum as the corresponding four-vector related to the space-time-translation symmetry of Minkowski space. This avoids a lot of unnecessary misunderstandings!