Is there evidence to support this? Short women have numerical and biological advantages (e.g., earlier puberty) wrt to pairing and reproduction; but aesthetic preferences seem tied to ratios—waist-hip, leg-torso—that can hold for samples tall or short. Heels seem like a means for adjusting the observed leg-torso ratio of a sample to mask suboptimal genetic inputs and environmental effects; i.e., heels make legs look longer in proportion to overall body height.
However, Choo and Ferragamo provide both offensive first-order and defensive second-order improvements to a given shoe-chooser's strategy.
Well-designed shoes enhance leg-torso-proportionality without triggering the "Hollywood & Vine" effect in the target.
Shoe-chooser cohorts (competitive or collaborative) incorporate their analyses of the chooser's tradeoffs into their feedback and behavior. Expressions of peer support ("props" in the vernacular) or criticism ("dis" in the patois of the street) respectively enhance or degrade the chooser's perceived desirability. A designer pedigree increases the probability of a "dis" being countered by a "prop", mitigating the Ho-Vine effect and further increasing the chances of the shoe-chooser successfully reproducing at her sub-optimal leg-torso-proportion.
Personally, I think women should stay away from high heels - they also ruin their feet doing it, which is definitely not aesthetic.