A "rule of thumb" in development of a product is that it takes...

A "rule of thumb" in development of a product is that it takes roughly ten times the cost, effort and timescale than to develop a prototype; most don't make it!

Dan Poneman highlights a DoE document (see https://fissilematerials.org/library/doe15b.pdf) from 2016 which (at page 35) says:

"SeparationofIsotopesbyLaser Excitation (SILEX)

Another potentially promising laser-based enrichment method is the SeparationofIsotopes byLaser Excitation{SILEX).Global Laser Enrichmentisdeveloping this technology but has recentlyindefinitely postponed planstobuild a commercialSILEXfacility intheU.S. Thisisa proprietarytechnology developed, in part, byanAustralian company. Therefore,its usetoproduceunobligated materialiscurrently considered prohibited based ontheagreement between theUnited States and Australia Concerning PeacefulUsesofNuclear Energy entered into pursuanttosection 123of theAtomic Energy Actof1954.
• Dueto the proprietary natureofthis technology, theTRLisdifficulttodetermine, but itis estimatedtobe 3. Global Laser Enrichment had suspended its planned developmentand demonstration program.
• Thereisinsufficient informationtoestimate the capitaloroperating costs.
•Ifa significant development effortis undertaken, then possibly this technology could becommercialized.
•Its usetoproduce unobligated material is currently considered prohibited based on aSection 123 Agreement.
• Additional scient ifi c and process development work would be required to producehigher assay enrichments.
• There is high technology risk and high schedule/cost risk based on experience with AVLISdevelopment, which took approximately 15 yearstomove from aTRLoftwo-threetosix"

This DoE document concluded
:
"SILEXisnota viable option under ex isting international agreements. In any case, thisoption would require a substantial demonstrationeffort forthe potential rewardsofahigh efficiency process."

In conclusion in 2016 the DoE practically wrote Silex off - we have much to thank the BoD (and Cameco) for their tenacity in hanging on. In the event Centrus built a 120 unit cascade of the AC100 centrifuge in 2015 which was scrapped by the DoE (who funded it) in 2016 and was much criticised for "wasting" US$400m of US tax payers money on this development since at that time (after Fukushima) nuclear energy was thought to be dead end. The cost of a Centrus enrichment facility at Piketon (400,000 S@WU/yr with 1,440 centrifuges) was put at:

"Assuming full fundingforthisoption , the first train could begin production in 2022 and be producing one reactor reloadworthofEUevery 15 months. The national security plant could be fully operational by 2025.Total costforthis option is estimatedtobe between $3.1 billion and $11.3 billion." - this is in 2016 US$ which is considerably more now. It also concluded that it would take 27 month FEED study and be in production by 2025

From the above Silex was practically written off by the DoE in 2016.

What we can thank is that the Piketon project was scrapped in 2016 and that Silex (and Cameco) carried on development of the laser (the highest risk component) and perfected this development with the Beam Control System (BCS) which was subject to 8 months of testing by Cameco and has now been delivered to Wilmington and a second laser is also now being shipped. I'm placing my faith in that Cameco will have thoroughly tested this component which is critical to the TRL/6 trial next year. I'm betting that this TRL/6 trial be a success and that the DoE have already backed it by endorsing the Paducah tails agreement and will want to "hedge their bets" on a technology diversification in backing both Centrus (for military) and GLE (for commercial enrichment with the hope that the capex for plants at both Piketon and Wilmington will be less that the previously estimated amount and will preserve competition to Centrus which has a poor history (as USEC). As I see it what Dan Ponemen says is correct but that many of the risks are now mitigated but this still depends upon the TRL/6 demo next year and Cameco's solid backing.

Taking your point on the ZX Si demo - yes I think this has done a lot to mitigate the risk but the laser has a different frequency which is much more difficult for uranium than silicon - but that the Cameco 8 months of testing (and the BCS) has to a large extent covered that risk - but we are not home and dry yet. What is important is not to be distracted by "rabbits" and keep on focus.