A lot of Gingery's stuff is obsolete in terms of not being 'expensive'. The books were based around the fact that scrap/metal were pretty cheap around that time.

Nowadays you're better of surfing Craiglist or buying an Import machine for most of his builds.

Metal is still cheap compared to new equipment, though not by the same margin. Much of Gingery's machines use cast aluminum, which is something that can often be found for free with enough patience - melt down soda cans, broken transmission housings if you have a mechanic friend, old furniture if you can find an office going out of business, etc.

Even at the time, you were probably better off buying a machine if you wanted one. However the learning experience of making your own is very valuable. The deep understanding of how the machines work, combined with practice means you are more likely to be able to use them correctly once you have one. (if you finish of course, most people don't)

While I wouldn't recommend a gingery lathe, I don't see any reason to ignore an injection molding machine design that is small, known working, and simple to the point of being obvious. You can certainly build one faster than you could find a castoff on Craigslist, and unless your shop is quite large you may wish you'd built the smaller machine even once you got your used industrial one working.

But would you want one as a hobbyist? Injection molding shines when it comes to mass production. For hobbyist purposes 3D printing seems a much better alternative. There are numerous companies out there offering this in a higher quality than FDM.

It really depends how you define “hobbyist.” I’m a designer by training and I (mostly) always designed purely digital products, but after moving to PM and consulting for a couple IoT companies I’ve become very interested in industrial design and design for manufacturing.

I’m especially intrigued by next-gen manufacturing (automated/lights out, 3D metal, etc.), and would absolutely love to have a small 5-axis CNC mill and some kind of injection molding facility alongside the 3D printing capability I already have for prototyping. Each tool enables a different range of options for design and material.

This isn’t now and likely won’t be my profession any time soon, so I would very much consider myself a hobbyist in this space.

Check out https://www.pocketnc.com/ - desktop 5-axis CNC mills.

I've looked at these, but if I went down this road it would take a fairly large commitment to begin with (I'd have to rent a separate shop) so I'd rather have something I could grow into, maybe a Datron C5 or similar.

3D printing is slow and parts can be hard to produce due to support needs and warp. A low volume injection molding machine would definitely be useful, especially if the molds would last 10's to 100's of pieces.

Sure, for the same reason that I have a vertical mill in my basement. It'll never break even, but it'll always be fun to hold up a part and know that I, and I alone, made it.

I have a mill and a lathe. These are wonderful for one-offs. With an injection mold I first need to create a mold and then the object. I just don't see the point (apart form maki9ng hundreds of the same object). Why not create the object directly? Afaik injection molding has even more restrictions compared to subtractive shaping (I've seen amazing things done with the latter).

Making dozens or hundreds of the same object is the point. The example army men from my linked article is a good one-- I could probably make hundreds of them on my creality, but it would take less time and the end product would be smoother injection molding them.

Perhaps as a rapid prototyping machine, this might work. As a hobbyist, I am not so sure.

The mill needed to create the mold exceeds your 2% pricetag.