Of course, they see it the other way around, that it's the non-conservative side of politics that made it polarising. That's the nature of polarity ;)
The adjustments to the historical temperature record are an under-discussed topic IMO. Most people don't know they occur. The justifications are by the way not that thermometers were inaccurately calibrated in the 1930s. Thermometer technology is very old. If there are datasets being adjusted that way, that's a new reason that I haven't seen when reviewing this controversy myself. The justifications are, at least for the USA, more like:
1. A claim that people before the 1980s didn't know how to use min/max thermometers and read them/reset them at the wrong time of day. So climatologists have a thing called time of observation bias adjustment, which supposedly corrects for this.
2. Thermometers got moved around over time, so they apply 'homogenization' algorithms to invent readings for thermometer locations that no longer exist
3. Thermometers often had gaps in their readings, so again, they interpolate readings that 'should' have happened from nearby thermometers. Because the thermometer network is getting older with time, the amount of interpolated data keeps going up. For instance the standard US thermometer network used for all American temperature timeseries now has >60% interpolated data generated by a model, they're not raw readings, because there's been so much movement over time.
The counter-arguments are:
1. It's unscientific to quietly alter your data to fit your theories. If there are doubts about the accuracy of the readings, you are meant to add error bars and incorporate that into your uncertainty, not just publish new versions of historical timeseries like climatologists do.
2. The historical justification for TOB existing is not very strong. There's an assumption that people back then didn't know how to use min/max thermometers correctly, but there's not much in the historical record to actually support this claim, which is an odd one because, again, thermometers are an old technology and people were making records of temperature for a long time. A min/max thermometer is not a complex device.
3. Putting aside the fact that again, supposedly raw scientific datasets aren't meant to have interpolated data points in them, homogenization might make sense in the USA where there was a large and well resourced thermometer network throughout the 20th century. But that's not true for the rest of the world, which was busy being wracked by multiple world wars, communist insurrections, 3rd world poverty and so on. Temperature data outside the USA for much of the 20th century is quite sparse and not very high quality, especially as people didn't really care about global warming until the 1970s/1980s.
4. Climatologists have continually adjusted raw temperature readings in new dataset versions even for very recent measurements, like measurements taken in the 21st century. These adjustments always strengthen the case for global warming, they never produce contradictory evidence. So there is a risk of a file-drawer problem, where now the precedent was set that adjusting historical data is accepted, people keep trying different adjustments with different explanations and end up discarding the ones that would reduce the perceived threat.