I've already posted a comment in response to one of the critical reviews, and it's almost as long as this review. I encourage anyone interested in more detail to also look there.
This book covers a lot of different subjects, but they all relate to the core subject, so that's a plus. The one reservation I do have is that I simply think Sivaram's timetable for the growth of solar is too conservative, and his assessment too pessimistic. What he thinks may take 80 years, I think may happen in 30, or less. His recommendations seem to make very good sense, although they get into a lot more detail in politics and economics than I understand. For that reason, I could not put this forth as an explicit criticism, or there might have been four stars.
I'm finding it a difficult call to decide how serious the obstacles are that Sivaram cites. He says that innovation needs to accelerate in three distinct areas, and he's probably right. But the fact that something hasn't happened yet doesn't necessarily mean it's not on the way to happening. In particular, it may not be the future development of solar that is at risk, but US leadership in this area—lamentable if true, but not as bad as it not happening at all.
I still don't really understand why Sivaram doesn't think energy can be totally based on renewable sources. Yes, I know about the Duck Curve and all that, and storage needs to improve radically. But it looks as if prospects for that are good. It's of crucial importance that there are several really good ways to address the intermittency problem. After reading the book, I still just don't get why a certain portion of fossil and/or nuclear needs to be retained, just because of intermittency. Between hydro, HVDC long-distance grids, demand management via smart grids, pumped storage, advanced batteries, supercapacitors, V2G, a bit of geothermal and hydrogen in places, and probably a few other things I can't think of right now, it seems we have a wealth of options without hanging onto the past, or going to molten salt reactors or fusion (although, if they happen, great, join the party, which is most assuredly NOT over).
In particular, I do think Sivaram exaggerates the problem of seasonal variation. If you think in terms of panels that track the Sun in both directions, there's as much sunlight available at high latitudes as there is at the Equator, on average. it's just that it varies seasonally. But when is all the heavy demand from air conditioning? In the summer months, just when the North has long days. There is plenty of flexibility in working this all out, though it will depend heavily on long-distance HVDC grids. (BTW I thought this book's treatment of the DC-vs-AC issue was, well, a ray of sunlight in the darkness, so kudos for that.)