Only 2 of 14 carb loading studies were properly blinded, and both found nothing

The authors evaluated 14 carbohydrate loading studies for blinding and ecological validity. Only two used a double blind placebo controlled design, and only four provided any carbohydrate during the exercise bout itself.

Both of the rigorous, blinded studies found no performance benefit from carbohydrate loading. The methodologically weaker studies reported moderate effects, at roughly the magnitude that would be expected from a placebo response.

The authors argue that once expectation is controlled and athletes fuel properly during the exercise itself, the case for carbohydrate loading becomes much thinner.

The numbers

Studies with double blind placebo controlled design	2 of 14
Studies providing carbohydrate during exercise	4 of 14
Performance benefit in the two blinded studies	none found

What this does not show

That carb loading does not work. The paper reports that the evidence is weak, not that loading definitively fails. Absence of evidence in two small trials is not evidence of absence.
That effects were pooled. This is a methodological evaluation and commentary, not a meta analysis. No effect sizes were statistically combined.
That in race fueling replaces loading. The reframing toward in exercise carbohydrate is the authors' argument, not a tested head to head comparison.

Caveats worth holding

Commentary, not a meta analysis. It reports that the evidence base is weak, not that loading does not work.
Absence of evidence in two small blinded trials is not the same as evidence of absence.
The appropriate framing is that the research behind carbohydrate loading is thinner than expected, not that loading is a myth.
Summarized from the abstract; the digest advises verifying against the full text before citing any claim.