The study seems incomplete without an understanding of what features the test group think the penguin should have (like the author described about height, beak etc.). I wonder if there would have been similarities that would have exposed either cultural or environment details about information bias.
All it seems to be saying is that mental archetypes differ in details.
> Imagine that, something as simple as a penguin, and take 100 people, and only 12 actually think they consist of the same weight, look, feel, and noise level.
The headline/title is misleading, implying that only 12% think that the picture shown is that of a penguin.
The end does give a glimpse of what we can to about it:
> Perhaps, before jumping at the temptation of one-upping another person with opposing penguinisms, we might try to stop and ask what the penguin did to them, first?
I take that to mean when encountering a disagreement, first try to clarify what parts you agree about, and convey (both ways) an accurate representation of the thing that's being discussed. I'm not even sure this is always possible, more often I find people talking past each other with unstated/oblivious implicits that are seemingly too obvious to even be aware of.
