I have been setting up the tooling I need to build an efficient, non-portable CO2 laser for the application of sintering glass. A portable version is difficult to ignore!
A rule-of-thumb says that you get up to 80W per meter of 100 torr sealed CO2 optical cavity, but with Q-switching this can be turned into high-intensity pulses. When you Q-switch a CW laser the CO2 takes a charge just like a capacitor and then when the stimulated emission part of the laser process is switched back on, the capacitor discharges all the energy it stored up when the q-switch was off.
Gallium Arsenide is an optimal choice for CO2 optics. Making lenses out of it isn't impossible, but it does require more than a passing interest. With a metamaterial of GaAs/AlAs you can exploit the berirefringent and piezoelectric properties of these materials to Q-switch at extreme Pulse Repetition Rates. This pulsed mode lets you make a very effective cutting laser since there isn't always a cloud of plasma in front of your ablation target, and the electrical requirements for the same effect are perhaps halved.
Pulsing can be taken to an even further extreme with Chirped Pulse Amplification.
Historically there has been little real effort to make CO2 lasers more effective than about 20% input to output. The length of the optical cavity for example is almost always opaque to mid-IR, meaning it requires a lot of cooling for all the energy it wastes as heat. The quantum rules of lasing however say that the direction that an electron emits a photon is always random, implying that the entire optical cavity should have its reflectivity q-switched. My guesstimate is that if you do this then the 20-80 ratio is reversed.
you're forgetting that to cut with a co2 laser, you need a lens. typically, the fl of lenses used for co2 cutting are 20mm, 50mm, or 75mm.
If you swap out the glass in an auto-focusing camera objective and slave it to another identical objective with original glass, you have a crude hack to focus the beam at your desired distance. This also lets you overcome the air ionization problem and you can even ablate mid-IR reflective materials by zapping the air right in front of them.
Air ionization means that quite a lot of energy gets reflected back into the lens and on the return path it may damage the system. With auto-focus distance data and GaAs/AlAs you could probably Q-switch again to occlude returns from re-entering the system.
