"Gbaike's solution is apparently diametrically opposed to that of Taylor."

Yeah, I guess that is fair to say:)

"A water payload has no intrinsic value unless and until it is in orbit, and even then, it presupposes a manufacturing plant in orbit capable of splitting and storing the H and O."

Actually, it doesn't presuppose. It does suppose the water payloads in orbit could arouse interest. The govt would be responsible for the water payloads, but would transfer ownership, if [at least] a portion of water was split and stored as H & O.
I think it would be in the best public interest that the orbital water was made into LH&LOX. But there could other uses of the water, such as using water as propellant [without converting into LH&LOX] or perhaps using water as part of some space station- sheilding, drinking water, even ballast type uses for tethers or whatever. I would say if the total was "looking like" it was only going to be 50 tons or less in one orbital location, then other uses could be considered, but the "highest bidder" would be the private entity which would convert the water into H & O [and would sell the rocket fuel or use it].

"There is no need to develop survivability for dumb loads of water. It's cheaper to just put another bucket on top of a new launcher. What you do push for is low launcher costs and/or reliability, since the cost of failure is the cost of the launcher times the probability of failure."

That's true, though one aspect of offering such low price was to somewhat discourage a launch company from solely depending on water payloads. Rather the idea is that as soon as the rocket proves itself, the rocket could demand a higher price for satellite launches. The more launch failures a rocket had, would require more successful launches to counter those failures. If for instance, the govt decided to pay signficantly more per lb- say around 1/2 of the current payload price (around $1000-2000 per lb) then companies might only want to focus solely on delivering water payloads, and this would get more into an economy of scale type situation (as Robert mentioned earlier in thread) and of course, would more than double the total govermental program cost- which might also cause the govt to impose various rules to limit the use of the program.