not so much water on the brain?
Gbaike's solution is apparently diametrically opposed to that of Taylor.
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.
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.
Interestingly, however, if this programme were combined with a simultaneous push for a survival package for alternate expensive payloads on the same cheaply reliable launcher, then both schemes *could* be integrated.