I would be inclined to place an order for 3 to 5 units if I knew roughly when you expected to have chips ready to go, and if the units had all the accompanying accessories to make them functional miners, interface, powersupply, heatsink, case, etc.
As much as I like the idea of them selling a fully functional and assembled miner I would think that the absolutely most viable and smartest move would be to sell 'Moderate level DYI packages', as that would save on assembly time, manpower etc.
By 'Moderate level DYI package' I refer to:
A DYI package consisting of all the components for customers to build a complete miner, aking to building your own desktop computer.
The package would consist of all the components required including:
- Hashing card/cards with soldered ASIC and connectors
- Interface
- Heatsink (Unless standard heatsinks available from Newegg etc. could be used)
Customers could then supply their own DYI case, standard PSU and possibly heatsink.
This would allow a cash strapped and possibly 'under manned' start-up to deliver a working mining product without the lead time, investment and manpower required for assembly, testing etc. allowing them to reach market far faster then delivering a completely assembled and designed mining unit.
In my mind, bitaxe.ca aiming at becoming a 'Radioshack' of mining equipment would have a far higher chance at success then just copying the business model of what is now 10's of companies designing complete miners based on the Avalon reference design. Remember people, time-to-market is everything for BTC mining. And the audience is more than just a little tech savvy!
Just my 2 cents from 15 years of start-up experience.
Edit; In my statement above I am making more than a few assumptions aroudn bitaxe.ca based on avaialable information. If possible another route that I think would be exceptionally attractive would be creating a passive or active heatsink design around the 10-chip cards and delivering them as a '2.0 of the USB block eruptors'. Obviously requiring external power source, but still a very attractive proposition since there are available designs for 1-chip USB miners already and I am guessing that expanding this board design for 10 chips could be fairly simple.