I don't think there's any perfect solution - this is something I've been thinking about for years now.
The inherent problem with auctions is that, aside from marketing them well, the seller has no control over how the market perceives the value. What if it was a piece that was way more difficult and took more time, but was slightly different and not really in demand as much? It could end up selling for a fraction of what the maker actually wanted, and that's just one of the terrifying unknowns of running this sort of business. How do you project any sort of financial stability and actually carry through a solid business plan? Do you auction them on a regular basis? Seems like customers would get tired of that as well.
First-come-first-serve sounds like the most fair, but there's the problem with internet speed, timing, and time zone. There are makers that do email newsletters (which can be faster than loading up a cart and filling it out) for one-off pieces and the first committed response gets the knife, but emails aren't accurate within the minute. Sales absolutely get made in that minute, easily. Do you just pick the first response received, even though it might not actually be the first one sent?
Lotteries seem like the least troublesome but also the least pleasing.
Custom lists are too limiting.
Raffles are legal nightmares.