As a heuristic, any study, that is not a large, randomized, double blind, placebo controlled study that is featured in main stream media, especially if it is not a human study, is more likely to be wrong than a study that is not featured in the media.
This is because of Bayesian priors. To be interesting enough to be featured in mainstream media, it needs to be surprising or unexpected in some way. However, that means that the Bayesian prior probability is low, and thus, especially since p = 0.05 is used for publication regardless of how unexpected the result is, the conditional probability is lower compared to a study that was not featured in media (because it was not a surprising result - meaning that the prior probability was higher).
In this case, I think the idea was chosen for media publication not because it was surprising but because it appeals to our senses. "We" (the collective 'we') like the idea that nature is a network of things. "We" also like the idea that mushrooms are some type of central nervous system for nature.
This is because of Bayesian priors. To be interesting enough to be featured in mainstream media, it needs to be surprising or unexpected in some way. However, that means that the Bayesian prior probability is low, and thus, especially since p = 0.05 is used for publication regardless of how unexpected the result is, the conditional probability is lower compared to a study that was not featured in media (because it was not a surprising result - meaning that the prior probability was higher).