I read your "how this is different than a spreadsheet" blog, and none of the examples struck me as compelling, even though I use my Excel/SQL toolchain every day. Given how strong the ecosystems are around those tools, I think it would be very challenging to convert power users without some really obvious hook.
That one hook COULD be a good mobile client. For example, I have a dashboard built in Excel with an ODBC connection to a few of my company's DBs. However, I obviously can't see this on my phone, since no mobile Excel client supports ODBC (not to mention the macros I have built in).
I could maybe see a "disruption" angle, where you provide 75% of the functionality of Excel+ODBC+SQL+Workbench in one easier to use package, but I wonder how large the market is for people who want to work with that level of data but don't want power tools. I guess that's one area where you'd have to trust your market research/gut.
Could be wrong though -- would be great if something like this could improve the data/analytical literacy of the general population.
A) people don't really know they need db-like tools.
B) even if they did, setting up a database to have even a portion of the "up and running" aspect of Excel is extremely time consuming.
The closest I've seen is some companies that get advanced enough set up a database engine somewhere, then set up Excel to connect to it as a datasource (ODBC or whatever). This still requires them to understand how to set up a database, secure it, and model table structures around their data, which is asking quite a lot from someone who really just wants to create a few short forms to collect data.
It would be quite a bit of trouble if Excel users want to implement access control on whether some users can read or write certain entries.
This more closely mirrors the workflow in my office at least, where only a handful of people create reports, but maybe 10-20x the number of people just want to consume it.