Part of the fun of the experiment is questioning the nominative determinism here. Did cultures pick a name like Baobab for that tree because of something this effect is showing or is just coincidence? We don't really know. Linguists have often wondered if there is a "universal grammar" underlying even our oldest languages that we don't know the full rules of, but coincidences and convergent evolution in our languages seem to tease could exist. Kiki and bouba are one of the more interesting experiments in this space to ever get repeatable results.