>invulnerable components.
That would solve most of the problems, actually.
A great deal of the problems with making powerful lasers is that they are horrendously inefficient, which means that they produce an ungodly amount of heat.
I think something like 20% efficiency is standard, which means you are dumping 80% of the input energy as heat into the laser components.
If you didn't have to worry about them melting, you could probably scale it up a great deal.
Not sure what kind of bottlenecks you would run into if you aren't counting component melting, cause I think that's the primary one.
Let's see, offhand I know about 6 types of lasers...
Solid, Gas, semiconductor, liquid, electron, and bomb pumped.
Solid lasers (i.e. ruby) use a solid core, a flash tube coil, a mirror and a partial mirror... the first thing to go would be your flash tube coil, followed by your mirrors, and then probably your ruby.
Gas lasers are simmilar to solid lasers, but they use a gas filled tube instead of a solid core, and again the flash tube would be the first thing to go.
Semiconductor lasers are more efficient, but their size limits potential applications (such as DVD player lasers and such)
Liquid lasers don't have a flash coil, but you instead get heating within the accumulation chamber, not to mention the problem with your optics (mirrors, lenses) melting.
Electron lasers use electron beams and synchrotron radiation emission, and the primary point of failure is the magnets, IIRC, and the power density of the electron beam.
Bomb pumped lasers are however, Designed to overheat, spectacularly..... but they are nuclear munitions, so.... they also have limitations on their applicability.