It's also true if you want F = ma to continue to hold in special relativity, then m is a velocity-dependent matrix. Not if you describe things in terms of four

It's also true if you want F = ma to continue to hold in special relativity, then m is a velocity-dependent matrix. Not if you describe things in terms of four-forces and four-accelerations, which restores the simple proportionality. However, the spatial component of 4-acceleration is not necessarily parallel to 3-acceleration.