Sorry to keep jumping you on nuclear specific subject matter. You do a very good job of remaining as objective as possible, so please accept my props for that.
Barnard is a good researcher and convincing author and does have his biases (which to his credit he does his best to disclose), but he
is a bully and does need to signal his allegiance to the 100% renewables crowd to maintain access to the group broadly, especially academics. Take the conclusions he and his in-group members (Jacobson, Sovacool, Ramana, etc) present with a HUGE grain of salt.
These nuclear opponents aren't wrong in some senses. The market for gigawatt scale electric plants is extremely limited, and these things could be practically immortal if you refurbished them sensibly along the way. NPPs are insanely productive. There is a lot of excess to hand out over it's lifetime and the pigs have long figured out how to line up at the trough. You correctly point out that those in or adjacent to the nuclear industry (even in opposition) have figured out way too many clever ways to make money that does not involve building more capacity. This is a problem not exclusive to nuclear, but acutely expressed in US nuclear.
Vogtle was brutal for so many reasons. I don't know why the AEC industry and those who give FID to megaprojects continue to follow a process that allows site work to begin before detailed engineering is complete. Vogtle was a new design that started work too soon, with a trade and supply chain that hadn't built anything meaningful for decades, and ended up doing the complex concrete, cable and module work essentially twice. Of course a project is going to look crap when this happens. True in any industry, worse here because you get to charge premiums for nuclear grade works because... reasons? But guess what, SO MANY participants made serious bank doing this. No lessons will be learned, sadly. But is it actually "too slow and expensive"?
Yes, the 11 years and 17B numbers at Vogtle look like a pair of big, scary cherries on top of a poo sundae of experience curve data in North America. It is easy to point to and say "wow, too slow and so expensive/unaffordable".
But these data points are not compared to other jurisdictions or points in time where nuclear buildouts did work, or presented with context as to why there were overruns, nor contrasted against the projected benefits nor analysed from the perspective of total delivered cost to the consumer! People are innumerate, and it gets worse as the figures get larger.
For Vogtle, Georgians will pay $75/MWh for the first few decades and $25/MWh after that. Is 2.5 - 7.5c/kWh "too expensive" for the average person in Georgia in 30 years from now? Fat f'ing chance. We pay more than that right now in Alberta, one of the lowest cost power generation bases in the world. How does that work?
Not only that, Vogtle's 4 pack will run for 80-100 years+ and will produce as much economic output as a small rich nation's supply of power over it's lifetime.
It will ensure that Georgia will continue to be able to attract and subsidize major factories in appliances, petrochemicals and pharmaceuticals away from places like Alberta.
It will provide multiple generations with very well-paying jobs in multiple areas of high value add skills, and is fully capable of funding pension obligations to union workers. Etc, etc etc.
How can one look at the entire picture of benefits stacked against an outlay of 17B that is rapidly being inflated away and say "that's a bad investment"? The ONLY way is by having interests that are not aligned with the typical consumer. That's it.
The real problem is that the story goes well beyond the scope of any economic analysis that we can conduct today. Our "economics" attribute zero value to anything more than a few decades out at best. "It's too expensive" is a near meaningless protest and should make you question from what perspective is that statement being made before believing who utters it.
Check out
Lovering et al (2016) for an updated look at global experience curves. This does not include UAE Barakah which was a stunning success by nearly every measure to supplement the nice OSTI paper you shared. Nuclear is not expensive everywhere, but it does take forceful owners to ensure that the benefits of production are not left to unabated lechery. It typically also requires institutional involvement with a commitment to long time horizons and drivers outside of maximizing NPV/next quarter's earnings per share, which is antithetical to the Age of Narcissistic Individualism we currently find ourselves within.
For more Canadian context, check out C4NE's
"Case For Candu" (2023) as Ontario is one of the few jurisdictions that would be well suited to build more GW scale capacity. Atkin Realis' rebranded EC9, the Monark (barf) could be a very good fit for a modern GW scale build out should Ontario commit to what is being discussed at planning levels wrt to expanding nuclear capacity, or is evident when analyzing forward supply-demand projections. It's plausible to see room for an additional 10GW in nuclear capacity in that Province, lest it be built out with gas.
I do tend to agree that SMRs are not going to be the silver bullet some promote them as. They do represent interesting tools if they ever make it to market. NuScale for example looks DOA, GE-Hitachi BWRX-300 looks much more promising. TerraPower's Natrium and Oklo's Aurora look better than ever now that Gates/Altman have been able to use the threat of AI's capabilities to convince Washington to increase access to HALEU. Many others are in the race with specific end users in mind... those who are not relying on random municipalities to invest in major civil works for the first time in 75 years (or ever) are going to do better than those who are still running that playbook. Municipalities are always awful customers, not sure that is news.
From a Canadian perspective that might look to contradict my position... OPG's foray into building out SMRs at Darlington is unique. As stated, I do not think that SMRs will be very successful in GW scale, electricity only markets, which is what the SMR project at Darlington is. But it is a sensible step to leverage an existing site license and parlay it into stepping out of the GTA as a business. Only a matter of time until the OPG-SaskPower JV is announced to build out a twin 4 pack in the Estevan area. They'll be fighting for their spot in the delivery queue as GE fulfils orders in Poland and elsewhere. By that time, Alberta might have a coherent stanza to add to the Canadian nuclear epic. It really should be addressing our incredible heat and energetic substitution demands which dwarf our electricity consumption, but... we will see.