Agreed, start with simple detection.

I've been wondering about a business case for asteroid data. In this case, an instrument would be placed in a halo orbit around the Earth-Moon L-1, looking outward. Over the course of each halo orbit it will scan a 360 degree loop of the celestial sphere, and over the course of a month it will map the ecliptic, especially sunward, which is where the true danger lies (in the form of ignorance of what's there).

While there would be an initial interest in the placement of NEOs throughout the Solar system from a few sources, if the person trying to sell the data wants to sell to more markets they're going to have to include more information. Spin rates, basic composition, and so on in addition to identification and orbital characteristics.

Businesses might sift through the data looking for targets of opportunity for exploitation. Scientists will want ever more data to answer scientific questions. Physicists will want to look for anomalies in the data that hint of unknown effects. The real question is does scientific data carry a realizable means to make a profit?

I'm not entirely sure it can, as the person supplying the data would be competing with NASA, which gives out data freely (relatively - the NSSDC does charge for copies of space data).

The subscription model offers an interesting approach for financial reasons. Each subscriber, be they individual, organization, not-for-profit, corporation, university, or governmental or inter-governmental organization would get a quarterly update of the data on the NEOs. By adding additional datasets from additional programs is a way to increase the cash flows from your subscribers.

I know of two good examples in the commercial world - WestLaw sells legal data, Bloomberg sells financial data. I want to sell space data.