I have no problem imagining renewable and storage taking care of baseload. Seasonal storage and winter peaks in cold climates and long distance flight are interesting problems. But meeting the baseload more economically than coal, gas or nuclear? Just roll out more of the tech we already have and which is already saving us money as we roll it out at smaller scales. Every small battery that replaces a horrendously expensive gas peaker plant is one step closer to the goal and generates savings we can spend on the next step.
10% is lighting which realistically can't - you need lights when it's dark.
15% is computers which can sort of load shift because most office computers are now laptops - with good communications systems you can tell laptops how aggressively to charge / whether to run off battery.
The rest is "other" quite a lot of which is small electronic devices like tablets and phones charging and then a grab-bag of misc. stuff.
Half of residential electricity is some kind of temperature shifting (heating, cooling, refrigeration, freezers, ventilation).
Essentially with good insulation you can get to a point where you can load shift 50%+ of your electricity consumption on an intra-day basis.
Electric car charging of course is a perfect load to match to renewables as well.
That doesn't solve the issue of inter-day fluctuations - what happens when you have a cloudy, high pressure system sitting over a substantial part of your generation for a week - but it does mean that we can go pretty far. We may have to keep more open-cycle gas around for those times which will have to be funded on a capacity basis.
Maybe in SV but not elsewhere. Plus servers that run 24x7 are probably the bulk of that 15%.
2) "base load" isn't the best term here. Base load refers to the load that makes up the bottom of the load curve, basically a straight horizontal line right under the trough in demand. Some of base load actually is flexible load - for instance refrigeration equipment that runs overnight can be timeshifted by an hour easily (although in that case it is rarely done because power prices are cheap at that time). The problem for renewables is not base load but non-flexible / non-dispatchable load that must go on when requested such as lights or televisions.
Renewable oversupply can already be economically stored for daily cycle purposes with 4 to 8 hour battery banks. If longer term storage is not developed, biomatter and synthetic fuels can run thermo-electric generators on the occasions that renewable supply falls short of demand.