I don't discuss politics in public with anyone unless I know them very well because the risk/reward analysis always falls into too much risk and no reward.
We discuss politics in the voting booth.
Twenty-five percent is not a small group.
The author pretends there is a huge majority (67%) who are sick of it.
There is no effort to bisect the “majority” into the readily known groups we know about. We know that 35% of voters are conservative and 21-25% are leftists. These percentages have held for over two decades and they almost always deliver.
35 plus 25 is 60 percent. So 40 percent “are sick of it.
40
35
25
Conservatives are in the second-largest group of three, more the a third of the country.
I agree with what you are saying—I would not dare discuss politics with strangers—it is too risky.
However, most people _really_ don’t want to discuss it, not _just_ because they don’t like potential snowflake emotional land mines, but also because they really don’t _like_ politics.
On a personal basis, most of the people I know don't care that much about politics except on a small number of issues that directly affect them personally, and even in those cases, they're "moderates" in the sense of having a mosaic of views: there are lots of people who are fiscally conservative but socially moderate or liberal, and there are lots of people who support a govt-funded social safety net while being social or religious conservatives.
So there's a lot going on that isn't captured by the choices on the ballot, and mostly consistent with the authors' observation that fewer than 1/10 of the population are Social Justice Warrior-types while only about 1/4 are committed across the board conservatives.