I don't see a conclusion here

About the only thing I agreed with was the idea that if you want people to use up their desire for tribalism, have them follow sports instead of politics.

I suppose it is possible to say that measuring pro-Republican or pro-Democratic politics correlates with some version of liberal-conservative axis. But it's a very strange version.

One because what constitutes "conservative" or "Republican" has varied widely even within the last 5 or 10 years, ditto for liberals and Democrats. Saying that these are indeed even correlated groups (that is that liberals are Democrats or Democrats are liberals) is relatively questionable as an assumption.

Second, because there is plenty of agreement between these two parties which goes ignored, particularly as it relates to media bias where the bias runs toward statism, regardless of its origin. Definitions of marginal terms are allowed to be skewed by their use by people have little in common with the real political ideologies that underlie them ("libertarians" and "realists" are prominent among these terms. As are "patriots".)

Meanwhile. The test used to generate quotients for instance tends to ask lots of biased questions that favor issues like technicalities and riders and amendments to overall bills without explaining the overall value of the intended bill. Opposition to these bills could have been based upon both the technical or overall value and little interest was taken to distinguish. The problem with this is that these technicalities are invoked by either political party to suit their pursuits. The timing of TARP related legislation for instance was crucial to determining whether or not it was a "conservative" division issue (ie, during the Bush administration, it was reasonably popular, under Obama it is almost universally reviled). As a related problem, almost none of these amendments and thus questions dealt with what are popularly defined as liberal-conservative divisions (gay marriage, abortion, welfare policies). The gaps between were not brightly defined and thus were putatively useless.

Finally, it is possible to account for "bias" that is observed or perceived by seeking alternative sources or balance and still finding the purported biased source to have at least occasionally valid uses to perform some other social function. It is also possible for that bias to be less easily measured or quantifiable as is proposed here. But in the case of the bias being measured, it is unclear that measuring it served much purpose. One basis for this would be the overwhelming appearance of American political figures to be mostly "conservative" in the global sense and the traditional sense of the ideology, and the purported appearance of mostly "liberal" political coverage and mostly liberal Democratic politicians. Bias in this sense would comport not to some structural ideological divide, for which it would be rendered useless to use any source of material for information, but rather for team sorting mechanisms. For which it has obvious intentions that can be easily sussed out. 

Suggesting "bias" as a result looks to be serving some other function, like tribal signaling, rather than serving some actual political purpose, be that nefarious or benign. A story or podcast about tribal signaling generally would have been far more interesting and factually based than the one they did. That is, they did the podcast on tribal signaling months ago, and that was a good one. This was bogus.