Through my years of work I have been involved with many different laboratory-grade photonic systems and engineering of such systems. Some work with the more common photonic systems such as magnification and prism optics, to advanced photonics such as beam-manipulation (beam correction/polarization/combining), color projection (RGB-mixing), high-output xenon-arc system tuning and projection, various laser power supplies, LEDs, and last but not least, lasers and many laser diodes.
The title I'm sure gives away where I'm heading with all this.
3D printers have begun to use a standard hot-end or replaceable extruder. I have worked with many different laser diodes, and I will start by saying that no current laser diode will cut anything thick while becoming a part of a 3D printer. However, laser diodes which can do some amazing things due to their power-density (not total power) have become abound. The requirement for a laser "extruder" (it is fed energy, and it "extrudes" many photons!), is not all that much power to do some amazing engraving work on various plastic substances. Two types of laser diodes exist (single-mode and multi-mode). Often the more powerful type is the least desirable for DIY engraving/cutting work (multi-mode output is not as concentrated and circular at the point of focus), but it really just depends on the intended application as to which diode will do a better job.
I have come up with a design which would allow integration of standard 5.6mm and 9mm laser diodes into a print-head as a hotend swap. The heat-sinking required is no more than a typical hot-end, when active fan-cooling is used, and something like the E3D V6 design replicates essentially what the laser head would look like when used. Power supply is simple and compact for these systems, as they are not made for cutting steel. The beam basically displaces the plastic it encounters. This occurs until the beam makes its way through. After the beam passes bed glass, it can easily be dispersed with a backing color material.
The laser function is similar to a fan in control, 30-100% on, and off. It is aimed perpendicular to the part down Z axis at 90° to surface, but could cut on any plastic surface within ~50mm of Z axis range or so. It should be able to cut through thinner plastics around 1.5mm thick, based on experiments. This is considering a steady moving speed.
These are the things which I can provide--the working essentials and fit. Someone else would have to know how to G-Code the laser movements. Would anyone be interested in such a project for a laser engraving or cutting system?