A more reasonable hypothesis: the severance of the autonomous afferences coming from the heart affects the brain stem and / or hypothalamus, which in turns affects the rest of the brain.
It has the advantage of being testable.
Another (IMO less likely) possibility would be that the lack of brain-induced rhythm variations of the denerved heart has an influence on the brain.
Palatability is an unnecessarily pejorative shorthand for agreement with one's expectations of reality.
I don't read "speculation" as meaning anything other than that the object of speculation is as yet unproven, so speculation that makes an argument based on previously known physical principles can still fairly be called speculation. However, such speculation can rightly be given greater weight than speculation which draws no connection to prior knowledge.
For example, if one were to argue that fairies make plants grow, with no other information, then their argument can safely be ignored. On the other hand, if their argument includes a testable definition of a fairy, suggests a mechanism by which fairies add matter to plants, and provides a way of verifying that mechanism, then the detailed fairy argument should be given higher weight than the unsupported fairy argument, until evidence is available that either disproves the definition of a fairy, disproves the mechanism by which fairies make plants grow, or provides greater evidence for a different mechanism by which plants grow (mitosis, perhaps).
Based on the wording of the abstract, so are the study's authors. (That is, they say "we suggest" as opposed to "we hypothesized and provide supporting evidence." But, maybe they do in the full article, I don't know.)
It has the advantage of being testable.
Another (IMO less likely) possibility would be that the lack of brain-induced rhythm variations of the denerved heart has an influence on the brain.