That's an interesting diagram, oren! The one thing that isn't intuitive for me, though, is why the initial longwave energy emitted from the surface is 110 units. Where does that come from?
Way outside my lane and just my understanding, but in case it's useful: It is counterintuitive to see 110 emitted when only 89 were supplied, and thus a -29 deficit. Everything has to balance in the end, so who makes up for those -29? My guess is that shortwave energy that reaches the surface, is absorbed and converted to longwave (infrared), and then reemitted upward from the surface, makes up that difference.
That means that the Earth receives more shortwave than it emits, and emits more longwave than it receives. The net energy has to balance, but within the total energy budget, energy from one wavelength bucket can translate into energy in another wavelength bucket, i.e. from shortwave to longwave.
A black tar road receives a lot of downward shortwave and a lesser amount of downward longwave radiation. The shortwave is absorbed and converted to longwave, and the tar road emits a lot of longwave back up (the road gets hot in the sunshine).
My objection to Walrus' statements is not so much about the unsupported statements about a specific mechanism that this discussion has evolved into, but conflating the "warmest temperatures of the year" to overall cooling, and using the comparison of 1986-2015 vs 1900--1960 observations to conclude that the current change is towards cooling, esp. when the projections looking forward show increased warming for all measures, including for warmest day of the year.
The original question was about why a specific location would be warming less than the rest of the planet. The Tamino article addresses that very question with his usual superb skill, including a discussion of and links to recent peer-reviewed studies for anybody who wants to go into it at depth. I assume that if Walrus' hypothesis had any credence, then the articles that investigated the southeastern U.S. "warmhole" would have included discussion of that as a factor, and that it would have thus shown up in Tamino's summary of the findings of those studies. But it does not appear in that discussion, which is not surprising because if it were true it would apply everywhere and would not be a localized regional influence.
I have not confirmed that assumption and not interested enough to do so. But if somebody wants to, that would one way to close the book on Walrus' hypothesis.