Mills discussed this in his lecture that I cited. The problem was achieving a power density that made it economically competitive.
Here's what Mills is quoted as saying back in 1991:
ie. a prototype CIHT cell that was producing 10 watts (not kilowatts) of power.
The contract never materialized. They were pursuing a contract, which didn't pan out because they weren't able to achieve adequate power density to make the unit economically competitive.
Mills is a little loose with the term "capacity" here. It wasn't putting out a 1000 watts. It had the potential to scale (capacity) to 1000 watts, but the 1000 watt prototype was never built.
The size and cost of the units were ultimately not competitive.
Consider you could buy a 1000 watt gas generator for a hundred bucks that burns 1 gal of fuel for 5 hours, using 1752 gallons over a year for 24/7 operation at 2 bucks a gallon, for $3500 a year in fuel costs.
For example, if a 1000 watt CIHT cell cost $35,000 to produce and install, that's like running a gas generator for 10 years anyways, and if the projected lifespan is 10 years, you may as well just buy the gas generator.
Of course, grid tied power is far cheaper than even a gas generator would be, so there was no point to going forward with a commercial variant of the CIHT cell.