If it falls much below that unit volume, the scheme implodes because the optional components will not find enough of a market to get manufactured.
In order to succeed, the internal connectivity scheme needs to not go obsolete over at least two, and preferably three product generations.
It also needs to find a market. People still build their own PCs for gaming, for lab automation, and other distinctly minority pursuits. So far, the niche for this product seems to be "people who find the idea attractive but can't articulate an actual need."