Don't care about the commercial argument. We have to do than spending time trying to sell free software and in particular for integration in another free component.
Regarding DRB, I am afraid it is the lowest level useful if you need an abstract access to all the supported types. I would sum it up as a sort of DOM or JDBC for more than XML or Databases, over which you can build software independent from the underlying data model and format (or at least at an higher level than usual): you should also know that it is a quite lightweight API that don't require depend on or learn more than 10 classes for many cases. If you are looking for a specific format, I would then recommend to use the dedicated API, if exists.
I don't know that much about the NetBeans Data Management System, if we are talking about the same component, but I know it is DBMS oriented which is not always suitable for Earth Observation data we had to cope with (the story of the integration of software between DBMS and XML is an example of the ambiguity for only two models when we are trying to achieve thousands). I don't say that DRB is a must but XQuery over data different from DB or XML as shown quite interesting results e.g. a single query for comparing or computing data extracted from all an ENVISAT MERIS, a Landsat and an IRS product, and not only for metadata because DRB does'nt know the difference with data.
Finally, DRB is currently has been used in ESA projects hidden behind higher level systems but has been designed to be open to other domains: we only exposed it as an independent and Open Source software quite recently. I hope that not all potential users will avoid it because it is not already used: otherwise not that many new stuff would emerge, but that could be the case. Whatever, we cannot do much better than distributing it as Open Source
Don't hesitate for other clarifications.
Stéphane
).