Building 4 hours of storage is a rather extreme step that's also extremely unlikely. California already has a 40c/kwh cost differential between peak summer demand and the middle of winter, but nobody is building massive grid energy storage to smooth that out. Even over the course of a single day you could make 10c/kwh just from grid storage but nobody is building it. Rather we add peaking power plants because base load power + storage costs more than peaking power plants and that's likely to continue to be the case even in a world dominated by solar + wind power plants.
A hydro plant running at baseload has a river downstream that is like... a river. Water levels vary with something close to the natural rhythm.
A hydro plant used for peaking will have no water flow for hours, and then heavy water flow. That puts extreme stress on organisms living downstream.
http://www.deseretnews.com/article/700006602/USGS-Gains-from...
Hydro here have carefully regulated flows to reduce the natural variation in the river so that for example an optimal flow is created for the various salmon runs.
Even at 4 hours of storage, the scale we need is a bit mind-boggling.
Your link is really interesting though. That's the kind of analysis we need. I see too many articles that just talk about rated capacity with no consideration of actual output or reliability.
My second link does similar analysis for Australia, and there it doesn't work out so well. The wind drops low over wide areas for days at a time.
In any case I have nothing against renewables, as long as they don't use up too much land...offshore wind is perfect. But I think we should be building GenIII and better nuclear reactors, too.