It's hard to assess performance and operation from a web-based forum. That you can test-fire the laser fire suggests a constant demand generates full optical power (and absence is off), and that suggests the laser diode itself is "working" (and for the supplier, that's probably the position they're aligned to). The only question I would have is the operation of the control circuit built into the laser assy - that allows the diode to be modulated with the TTL input. It **should** be a case that "on" is lase, "off" is don't-lase, and that is simply and exactly what an Arduino PWM output will produce. If the laser assy is questionable, then that can be assessed with passing a variety of signals into the laser and observing the resulting laser output. Similarly, you should be able (with equipment) to monitor the PWM output from the Arduino, that under GRBL it generates a pulse with variable mark-space ratio up to full-on (or near-as-dammit). Without testing these it's difficult to throw blame at either the laser or the arduino/GRBL.