Are Trump and Clinton ~Down~ with Downs's Median Voter Theorem?

This hilarious clip from Saturday Night Live's version of the Final Presidential Debate is remarkably close to what actually happened. Although it started off as a fairly 'normal' debate, with both candidates trying to portray themselves nicely to voters, shots were quickly fired. While I sat there watching the actual debate, I found myself infuriated by many things both sides had to say. I consider myself fairly close to what the median voter might be, so I wondered to myself if either of these candidates were successful in navigating themselves to capture the median voter from Downs's model?

This is an atypical election in that the voter population distribution is so polarized and bimodal that even the median voters are being alienated by candidates who are (supposedly) moving towards the median position! Hotelling would say that in any major election, the candidates function similarly to competitive firms, meaning that there are little -- if any -- distinguishing features between the two in an effort to gain the most votes. This could not be further from the truth this time around! Trump seems to alienate people by simply opening his mouth and pandering mostly to wealthy elites and white supremacists. But this makes at least a little bit of sense since he is not a career politician and therefore may not be well versed in the art of formulating a campaign that suits the voters' preferences instead of his own agenda (which we can debate over the values of this, just maybe in a less extreme candidate). Meanwhile Hillary Clinton, the career politician knows fully the importance of winning the median vote, but has flip-flopped so many times that she has ALSO alienated many voters! The SNL skit goes so far as to say: who do you want, Trump, or the Republican?! Trump is given his own category, meaning according to SNL he hasn't even come close to a platform supported by a median voter. Comparing Clinton to a republican is a testament to her attempts to gain more votes, but her more "moderate" platform may not be enough as she has already alienated people from her constant changes of pace, who's to say she won't change her mind again when she's sworn in?

These are all points that may come into conflict with the Downs's Median Voter Theorem, but it still is not enough to discredit it. If there exist two completely different candidates with voter distributions that do not even overlap AT ALL (which is slightly more extreme than even that of the 2016 election), then it will simply come down to how many voters there are in each of those distributions, and how many people either candidate has alienated within their own distribution. In other words, there will still be a winner, but it probably won't be one the true majority of Americans are genuinely happy about. (I wonder if we had another election system such as the Hare, Coombs, or Borda Count would the election look a bit different?)