If it's the boring corporate view, it'd simply be loosely packaged. Console pricing would still be kept as premium as possible, but as low as they can get some reasonable volume going.
Then they would simply roll out X camera-savvy games per year for the gaming population. From their macro perspective, they seem to only look at volume numbers (like a typical h/w manufacturer) and do some backroom calculation to project their growth. The projected number needs to be large enough (like one billion dollars) for them to take it seriously. This should allow them to control their cost/profitability in a familiar (and hence, comfortable) manner.
In this view, PS Eye was bundled with Eye of Judgment for $69. The added controller would push it higher but should not be more expensive than a complete Wiimote because part of the tracking mechanism is already in the PS Eye. If you already have a PS Eye, then the entry cost will of course be lower. The thing is the user can buy-in incrementally because with a PS Eye, you can already achieve part of the functions. This is the hardware centric, risk averse route.
If, by the chance that someone in Sony is in a startup mood, then they'd look at the target market (e.g., women), and then dream up a complete package to draw them in. e.g, Hands free UI may be useful for women with dirty hands (in the kitchen), so expect some apps for the kitchen; they may also create a Magic Mirror type applications to address their daily dressing needs; etc. Then, Sony has to figure out how they intend to charge these people. They could go for a subscription model, and use it to subsidize the hardware cost. In this manner, the BOM cost is not a showstopper, but still important. They could also do a razor model if there is a consumable somewhere in the plan.
So basically, everything is possible.
And they can technically execute both moves above.
Did Toshiba's demo not manage this? 
