you're correct about development cycles decreasing, however...

you're correct about development cycles decreasing, however economic realities will eventually define the minimum cycle length. Developing new technologies is incredibly resource (time, expertise and money) intensive and unless developers obtain a return on their expenditures they wont initiate new cycles. New tech must have a minimum time in market to be commercially viable.
With this in mind, its reasonable to assume that new tech must meet certain requirements to be considered for development including dramatic performance improvements, ease of manufacture (less new tech to manufacture the new tech), minimal disruption to existing manufacturing facilities if possible, an obvious need (market opportunity), etc.
RRam and WBT would seem seem to meet most of these conditions IMO. Performance improvements, simple manufacturing requirements with no exotic materials needed, BEOL, scalability, etc, etc. These are all the ingredients for a successful product.
There's still no guarantee that WBT's chip will dominate the market but I think it will carve out enough of the pie to be a success until the next advance comes along.