Overall, the experience was very good, we learned a ton of things, met interesting people, such as a guy who explained in details how to use that magnet printer, and some mechanical engineers building a cool sandwich-making robot which later turned out to be a YC company. We had a lot of fun.
However, as soon as our project was completed, we discontinued our membership. TechShop's pricing is weird: it's either insanely cheap if you're working on a project full time (I would have paid 2 times the amount no problem), or it's prohibitively expensive if you're just there 2-3 times a month.
Curious, was this a first BM project? (It's certainly up at the level of complexity that very often completely misses the deadline and either gots take out to the desert half-done, taken out a year later than planned, or abandoned altogether.)
I think the key is to choose your complexity. Some things may look complex, but they are actually just the juxtaposition of some simple, repeated elements.
Ironically (speaking as a software engineer), the one thing that we didn't finish on time was the software. We ended up running some simple code to power the colors, as opposed to the fancy fluid simulation that I had originally envisioned (http://gregschlom.com/flow-and-wonder/)