The experimental setup was basically:
1. Divide participants into builders or non-builders
2. Get the builders to assemble an IKEA box, and have the non-builders inspect an assembled box
3. Get the builders to bid on the box they assembled, and the non-builders to bid on the box they inspected
The results sound very impressive, with the builder group bidding 63% more than the non-builder group.
Looking closer, however, the absolute magnitude of the change is tiny, with the builders bidding $0.78 for their masterpiece, while the non-builders bid $0.48.
It's clear that reporting a 63% bid increase makes for a more marketable study than one that reports a 30 cent difference, but it's unclear if this effect scales with larger purchases.
I'm not familiar with the broader literature, but from my uninformed vantage, it seems possible that the effect declines into insignificance with larger purchases.
The paper also reports a second experiment with origami, and the difference is $0.05 vs $0.23, which at least is consistent with the hypothesis that the effect size will decline as prices increase.