When my child was born I spent my paternity leave getting started with Arduino, and ESP8266 devices. These days I've kinda stalled, but there's something fun about making the hardware and the software, and getting back to the limits I used to "enjoy" with limited RAM/flash-space.
Mostly my projects are put together for learning, then the parts are recycled, but there are a few projects dotted around the house that have stayed in-place for a couple of years now.
(Wireless temperature/humidity sensors, along with tram-departure information, etc.)
I ran into my own discouraging brick wall when trying to figure out how to fabricate metal control panels/enclosures and printing labels on them without expensive cutting equipment.
I'm not sure what you mean here. Since it's just KiCAD and then you can get boards for cheap, why is it hard to start the investment?
Drilling is a pain in the ass. The bits break easily and drilling accurately, even with a press, is hard. Milling makes this so much easier. Then you need to think about plated through holes, do you rivet? Do you just put some 30 gauge wire and solder?
Nowadays I find I rarely need to rush boards and Oshpark is so cheap it's ridiculous. There are cheaper companies (I've yet to try Dirty PCB), but you get 3 boards, ENIG and really good tolerances for free. Even rush shipping ($20 international fedex) is cheaper than getting it done here in the UK. The only competitor I've found is Ragworm, and they're both more expensive and you only get one board with HASL.
Unless you really want a board made now, I would say just get them made for you. The quality is so much better without spending tons of time. Of course sometimes you really do want a prototype instantly and if you want to make a weird shape - perhaps a really long rectangle that would cost a fortune to get made, it's worth having the kit in your house to do.
On the milling side, I'd be tempted just to buy a Spokeo mill which can apparently do reasonably good PCB milling. Costs a lot, but at least it'll do other things as well.
I've recently got access to a lab with some very expensive rapid protoptying kit (from LPKF), which I'm keen to try out. It can even do 3D PCBs laid onto plastic components.
https://www.mouser.com/ProductDetail/474-PRT-10811
Now I'm using the same barrel jacks, but with an old linksys wall wart.
That said, it's not an ideal situation, and I'll most likely build something like this in the not too distant future:
http://musicfromouterspace.com/analogsynth_new/WALLWARTSUPPL...
I started a deep dive in electronic music and synthesizers in the last couple of months, and have been thoroughly enjoying playing the productions units I own (Korg Monologue, Elektron Digitakt, Roland JV1080), but have even more so enjoyed the sort of free form effort of building my own instrument:
https://imgur.com/a/FZ6GFsI
That started out as a nightmare breadboarded voltage controlled oscillator, to a nicely breadboarded vco, to a rebuild on perf board, and then obviously I needed a perfboard power supply. But, once I wanted a mixer, I needed +/- 12v, not just 5v, so that led to a better iteration of the power supply, etc. It's a rabbit hole, but an immensely educational and fairly inexpensive one. I figure I'm about 3 months and a couple of YouTube tutorials away from my first Kicad developed oshpark printed PCB, and I have a ton of ideas for next steps, and prototypes in flight (CES3340 based VCO, Arduino MIDI control and CV translation, Lowpass/bandmiss filters, so on and so forth).
Back to the subject of the original post, this totally looks like Pennsylvania, and if it is and the engineer who built this is in the job market, and has any interest in working for a smaller, established (20 year) shop that lives in one of the automotive spaces doing a lot of cool custom hardware/software development, I'd happily accept a resume.