The article makes reference to this being expected from the physics, but I don't think it's particularly intuitive why it should be the case. Unless it's as simple as smaller swept areas needing higher tip speeds for the same lift? That would make the energy loss to drag worse for multirotors.

Higher aspect ratio blades are more efficient. It maximises disc area for a given mass of blade (linear-squared relationship). Blade is held straight due to tension not purely from torsional strength, so it can be made from a material with high tension strength like carbon fibre which is light. Also it positions the fastest moving blade area far out from the fuselage where there is less blown-drag.

The downside is that the tail rotor is purely loss.

Potentially a similar efficiency could be achieved with two large rotors, but then you end up needing to use cyclic roll to counter drag-induced cyclic yaw, or adding extra pitch mechanisms on each motor mount to make differential yaw.