My understanding (which could be wrong, or at least inexact, given that we've yet to see an evaluation example of the product) is that the Connect product will upload exclusively to the wl.com network, at least in the short-to-medium term.
Personally I think this may be a bit short-sighted. Certainly, there will be quite a chunk of users who will be content to use the simplest possible way of getting their data online from remote locations and the wl.com approach will be fine for them, especially if Davis manage to expand the wl.com features to include eg charting historical data (which shouldn't really be a big deal with the data in SQL databases, provided there's enough server horsepower available).
But then there will be other users who will want more control or to do things more cheaply - I'm guessing that Davis see the Connect plans as providing a significant income stream for the future, hence pricing by frequency of upload, which isn't really justified by differences in the amount of data transferred.
As for 'Purchasing an annual service plan', no sorry no further definitive pointers on this at present. I was guessing that Davis were going to negotiate contracts with various networks and take payment themselves but then obviously passing on the payment for data transfer to the networks. But even within one country this could get tricky. I really don't know, but I wouldn't be surprised if sorting out how best to approach this conundrum isn't part of the reason for the present delay.