"especially when the cost of build/deploy <= cost of comprehensive research"

you REALLY think so? I agree with your first few points, but I disagree on that last (quoted) bit entirely.

Building/deploying something of real quality (and potential viability) can be very costly.

The point of the app is to make that "cost of comprehensive research" negligible in comparison to what it is today. Either you pay a ton to do it right or train yourself forever on how to do it cheap (which is a cost of time). We want to lessen that burden exponentially.