I agree with much of what you say Jon. I've been looking at 3-d printing for some time now, didn't want to buy one, most specifically because I could never decide between a CNC mill or a "hot glue gun" printer.

There was one that fit the bill (well, very close) - $999 on Kickstarter would get you a 4-in-1 printer from a reputable Chinese firm; CNC mill head, hot melt (dual nozzle) head, 3W laser head. The "4" came from CNC, additive, laser PCB and laser etching (so 3 tools but 4 functions). It looked great and I was checking and deciding - I was going to decide before end of Aug. Then today they pulled the Kickstart plug. Something about someone quitting and security of the kickstart. (maybe).

https://www.kickstarter.com/projects/662229155/formaker-4-in-1-cnc-mill-laser-pcb-3-d-printer-all/comments

As to "what to make", I have a few ideas. For example, I have a weight bench with the weight stack. I bought it used, and some of the ABS plastic weight inserts were broken, and some missing (they act as cushions and guides for the bar when lowering weights). I called the manufacturer, but as it was 1990 built, they had zero supplies and even less documentation about that model. Yet the parts are a simple ABS ring with 'catches'. I am sure I could design one in a 3-d modelling program. So with a printer I could make as many as I need of this obsolete, cannot ever be purchased again parts.

I find many similar parts when doing stuff around here. Being able to print the part would be great. Also I'd investigate using PLA (or wax) as a building material  and use the "part" as a lost wax mold for casting with copper or pewter or other metal.

But my biggest problem with the current crop of glue gun printers is that the resolution is pretty much crap. They have come down from the original 1/8inch material, but it's still pretty coarse stuff - you can see the ridges on finished products  so much finishing is needed.

I also love the resin printers, but they are very pricey for what is a small volume (x by y by z sized) print.

I like the idea of 3D printing parts for obsolete machines - that's certainly something I plan to use mine for, when it comes (I have an ancient outboard, for instance, with a sacrificial steering guide designed to snap off if over-taxed, for which a spare was provided with the original motor, that is not made any more and that must have been used up at least 15-20 years ago, long before it came into my possession). Not exactly an educational use though!

I've seen some quite good resolution from some devices, though it depends a lot on material and speed of printing. Mostly, for extrusion printers, better resolution means slower printing which, given that even quite simple designs can run overnight, is not a good thing.  I found a very interesting summary of alternatives to PLA and ABS at The Creative’s 3D Printing Filament Guide: ABS vs PLA vs many new, innovative materials that I'm hoping to explore. Some particularly interesting ones like nylon or those incorporating wood or metal that might avoid the need to go for lost wax casting.

One possibility intrigues me - the means to replicate physical objects. I can see this being very useful for people in fields like archeology and history of design, as well as for fixing broken things, but I have a gut feeling there ought to be uses for this beyond the obvious.