I've built my Boston now using the parts I bought from JLCPCB (also, all my spares have now been snatched up by other users), and it's terrific. I won't defile Pylon's thread with any photos yet, because my 3D printed case hasn't arrived yet so it's currently just the naked keyboard sitting on a piece of cardboard! But I've soldered it (all hotswappable with millmax sockets - not cheap, but boy does it make life easier!), set up the QMK, and am using it happily. It's easily the best keyboard I've ever used, though that statement is a no-brainer, because keyboards are all about functionality for me, and the Boston objectively has the most functionality of any keyboard I've used. I've been excited for this keyboard for a long time now. I'd actually been working on my keycap layout, switch selection and QMK firmware for months already! (By the way, 1.25u keys work really nicely for the nav cluster trio). The Boston was my birthday present, and a very successful birthday present it has been indeed!
My main criticism of the design, and it's a fairly minor criticism, is that I wish that the gap between the top 2 rows and the rest was 0.25u instead of 0.5u. Many 100% keyboards use a 0.25u gap here, so an extra-wide 0.5u gap seems a pity on a keyboard that already contains an extra back row that must be reached. Actually, this turned out to be much less of a problem than I feared it would be, although I have long fingers so others may feel the gap more.
The placement of the 3 indicator LEDs seems a strange design decision to me. When I saw that one of them was directly above numlock and one was directly below scrolllock, I naturally assumed that those 2 LEDs would correspond to those keys, with the 3rd (leftmost, hence closest to the alpha keys...) LED corresponding to capslock. Yet the way they're setup in the default firmware (and physically marked on the PCB) is different: the LED next to numlock is for scrolllock, the LED below scrolllock is for capslock, and the leftmost LED is for numlock. This feels unintuitive to me. It's not a big deal though, as you can swap them around in QMK if it bothers you. I personally don't even use those LEDs, as I don't use the scrolllock function, I've eradicated all numlock functionality from my system (I got sick of certain programs always switching numlock without my permission so replaced it with a customised numlock-like layer in QMK that those programs can no longer touch), and I've programmed QMK to display the capslock status on the RGB light anyway. That RGB light is great. It's in a good spot where it's conveniently visible but unobtrusive. Apart from capslock, I also use it to show when certain layers are active or when a dynamic macro is being recorded.
Once everything added up, buying the parts at JLCPCB ended up more expensive than I initially expected. But the process of ordering the parts itself was painless, largely due to Pylon's excellent instructions. Pylon's thoughtfulness comes through in other areas, too, such as the helpful markings on the PCB. For example, little arrows guide you to where to put switches/stabs for the myriad variations (eg. 1.5u ctrl vs 1.25 ctrl, or 6.25u spacebar vs 7u spacebar), which makes the installation process less error-prone.
JLCPCB is/was sold out of the MCU, so I had to buy the MCU separately on Aliexpress and drag-solder it in myself. To others who are new to drag-soldering MCUs like I was, my advice is to
make sure you have good magnification! I wrecked my first MCU because I
thought I could adequately see what I was doing, but I was wrong. On my first attempt, I used a desk-clamp magnifier that's more than enough for soldering switches, diodes etc. But the MCU is so tiny that it was still pretty hard to line up the pins correctly and spot soldering errors. After reworking the soldering job too many times, I ended up bending the pins all over the place and eventually snapping one off. So I had to order another MCU, and this time I also bought a better magnifier. Nothing fancy - just
this $30 one. It only provides slightly more magnification than what I already had, but that slight increase made all the difference. Lining up the second MCU was a breeze, and I immediately realised how limited my visibility had been the first time. When you can properly see what you're doing, drag-soldering is pretty easy, with the flux doing most of the work for you. Remember to use a very light touch lest you bend the pins.
I'm finding the layout to be great. As I mentioned, I'm using 1.25u keys for the nav cluster trio, which I've mapped to HOME, END, and DEL. Not only do the 1.25u keys fit very well aesthetically in this area, but the increased size makes them very comfortable to hit with your pinky. I use DEL a lot, and using a 1.25u version feels like this is how big it should always have been. The little 0.25u gap around the arrow keys is excellent too. I was initially sceptical whether such a small gap would isolate the arrow keys sufficiently (compared to a 100% style keyboard where the arrows live in their own separate spot), but I'm finding the 0.25u gap to be perfectly adequate. My fingers gravitate to and locate the arrows effortlessly, even though my muscle memory has never experienced this exact layout before.
Thanks again Pylon for this awesome workhorse of a keyboard, and I hope your arm heals quickly!