[Monochron]

I'm going to take a different approach to this and say that you are actually asking the wrong question. Rather than asking which is objectively better, you should be asking which drive is worth the risk vs. reward.

I think nearly everyone will tell you that swerve is going to perform better than mecanum in nearly every situation, assuming you can build and program it well. And I'm not just talking about precision here, using proper materials, being willing to hunt down ghosts in the code, and ensuring that your drivers have a huge amount of practice time is necessary to not lose control of the robot.

I have never been on a team that used one in competition, but have experimented with them in the off season. If your encoders telling you wheel position start to slip you end up with wheels at all angles. And thus, your robot may very well end up trying to drive itself in 4 different directions at once.

That being said, with enough practice at it, building a reliable swerve drive can be done as long as you can precision machine the parts for it.