The economic of new nukes (even "the little nuke that could" types) just don't cut it in the real world, just as Solyndra didn't/couldn't.
Large in Solyndra's failure was that they didn't think the cost of silicon PV could crash as quickly as it did, and they had little hope of getting their expensive panel/cheap racking approach to market quickly & cheaply enough. The panels used to be the cost-driver of small scale PV followed in short-order by the racking systems necessary for mounting them. Solyndra had bet that getting the cost out of the racks was the quickest way to cutting the cost of the system, but that proved not to be the case, and panelized silicon-PV was destined to stay ahead of them on the cost curves.
It's doubtful given the current high cost of building nukes that they would become even REMOTELY competitive with grid-hardened PV/wind and natural gas fired micro-cogens in any reasonable time frame.
And the Solyndra deal was a loan-guarantee, not a handout. In the end it won't end up costing the taxpayer the whole 585 megabucks. That loan guarantee program as a whole had already budgeted in far more than the Solyndra costs, and other US CIGS technology solar companies with very low-production-cost panels are in high double-digit expansion mode (notably NanoSolar, another beneficiary of the loan program, who maintains the bulk of their cheap printing-process manufacturing in California.) The US bankruptcy system is a slow greedy piece o' crap compared to how it works in Canada, but despite that the lawyers won't get away with vacuuming up 100% of the asset value on this one.
But one bad bet is by no means the signature of the US solar industry as a whole. If it's not developed here we'll be buying it entirely from China (and loving it.) At the moment the solar balance of trade with China favors the US, but it's primarily due to US exports of silicon production technology. Cheap Chinese silcon PV is driving some US PV manufacturers to pick up an move production there (notably Evergreen.) There's a lot of consolidation going on in the industry as the prices race to the bottom, and not all novel approaches to cost competitiveness (like Solyndra) will succeed- count on it.
At current pricing the lifecycle cost of power generated by silicon PV is already well below 10cents/kwh even for small (1-10KW peak) systems, and with CIGS technology in large arrays it's even less. Most analysts are calculating PV kwh will be at grid-parity with the cheapest fossil plants before 2020, maybe by 2015. There's plenty of room for PV taking decent slice of the pie by 2030 despite the fact that it's only making kwh-hay while the sun shines. With electrification of the automotive fleet and with smart grids & batteries it could run the whole show by 2050, but without the policy initiative to do that nobody is making that bet just yet.
Nukes (even mini-nukes) have the ramp-up/ramp-down time problem, over and above the daunting initial up-front costs. Ain't nobody expecting sub-15-cent lifecycle kwh out of any newly constructed nukes, even if maintaining and extending the life of the existing fleet is reasonably competitive with the fossil grid. If you can't ramp it up fast, you can't crank it back overnight- most existing nukes are heat-dumping at night when there is insufficient load to be able to meet the morning load. In heavily nuclear France they can sell some of that nightly excess to Spain & Italy (at a financial loss) to avoid cooking all the fish in the rivers, but that wouldn't be the case in a high-nuke USA. (Maybe an electrified auto fleet could soak up some of that off-peak excess, if that became policy.)
Even if it were burning raw crude oil at an estimated 2015 price of $100/bbl, at ~15% efficiency the dirty-little-diesel will not be competitive on a kwh cost basis with PV by 2015. And as a heating system a mini-split with an annual average COP of 2.5 with 5-cent PV power looks pretty good relative to the 60-65% heat output of that type of cogen with 100bbl oil.
Then there's the first ~30% of end-use efficiency improvements for that are essentially cost-negative- they pay off in utility bill savings in very short years. From a grid policy point of view it's cheaper to reduce the base load by a nukes-worth of pre-1990 refrigerators with higher efficiency versions than it is to build new capacity (any source) to support the difference, and that's just the tip of the efficiency iceberg. Getting rid of the worst-efficiency incandescent lighting technology might be the frost on the tip of the efficiency iceberg.
That's cost-negative and very cheap up front, yet people bitch about that. Imagine the hue & cry of surrounding them with mini-nukes (of any design) post-tsunami, and present them with the bill for it? The meltdown fear may be as irrational as the efficient-lighting nay-saying, but the pain in the pocketbook would be real. The nuclear industry didn't crash in the early '80s due to all of the protesters, it crashed due to the high cost, the WPPS bond failure, and the very real prospect of much cheaper efficiency & conservation cutting into demand- there was not a profitable market for the output of those projects, and it was beginning to be obvious. Building giga-projects with decade-plus construction schedules is just too financially risky with the amount of low-hanging efficiency fruit that abounds, and those factors haven't changed much in the past 3 decades.