Thanks for the info guys. No matter what I do for manufacturing (whether it be getting them professionally made from start to finish, or just getting professional PCBs and soldering them myself), I would be doing all testing/support/rework/etc myself. I don't at all think that the device I have in mind has a big enough market to require that to be contracted out professionally as well. The target customers are a bunch of DIY'ers and from what I have seen of their reception of similar products marketed by individuals or very small businesses (many of which have proven to have some issues), are a pretty forgiving bunch, so I don't have to worry about making a 100% bulletproof commercial product.
The typical process of release for such products has often involved a beta test phase, where 10 or 20 units are given out (or sold at a low price, perhaps at cost) to respected members of the community, who stress test them, report any issues, and write up some reviews if everything works well; that takes a lot of the stress out of it, not having to over-engineer the crap out of it first. If this whole idea of mine works out in the end, I plan to do this beta-test type of thing, to make sure my device is suitable for general release, documentation is complete enough, and it creates a foundation of (hopefully) good reviews.