Skepticism about the volume of power requests has a mathematical consequence that tends to get lost. Even accepting a 90 percent attrition rate on those requests, the ones that survive represent 47 gigawatts of demand. The conservative scenario is already embedded in that number.
The 47-gigawatt floor
Work through the arithmetic. If 10 percent of requests are real and that 10 percent equals 47 gigawatts, the full pipeline implies 470 gigawatts. Most of that will not be built. Attrition will thin the queue substantially. The point is that the thinned queue, on this logic, still produces 47 gigawatts of load.
That reframes the argument. The case against large demand projections has rested on the claim that most requests are speculative and will never clear. This logic accepts that premise entirely. The 90 percent failure rate is the concession, not the punchline.
The counterargument
The strongest objection is that the 10 percent survival rate is assumed, not verified. No completed vetting process has confirmed that one in ten requests is financeable, sited, and contracted. The 47-gigawatt figure rests on a filter that has not yet run.
A second objection follows. The framing applies a uniform 10 percent figure across an entire pipeline. Real attrition varies by geography and project stage. If aggregate attrition exceeds 90 percent, the floor drops. The number describes a scenario. It is not a measured outcome, and treating it as a planning anchor carries real risk.
On balance
The read-through is this: the skeptical view is already baked into the premise. Critics have pointed to the volume of speculative requests as evidence that demand projections are inflated. This framing concedes the point. Ninety percent of requests are treated as noise. What remains after that concession is 47 gigawatts. The line to watch is whether generation and transmission capacity can be contracted to meet a number that is, by this logic, already the bear case.