If you know more than other people, the thing to do here is neutrally provide correct information, so people can learn. If you can't or don't want to do that, please don't post.
If you want to convince others, you can provide a direct argument with data/popularly cited paper that a significant proportion the science in the field is bad. There are always reasonable people, not necessarily the ones you are replying to, that you may yet convince.
Alternatively, you can also provide an example for illustration and good reasons why it should generalize.
> If I mocked someone for using a linear regression on cyclical data, do you know why that's bad?
It's not necessarily bad to use linear regression on periodic data. If you wish to capture a trend independent of the periodic change, all the data in the periods are uniformly sampled/weighted, and the period is so short v. the timeframe of the data/the data for the regression is a whole integer.
If you calculate a linear trend from subsets and get largely different results, a linear trend though the entire data set is only meaningful if you have several cycles of data. I would argue we do not have several cycles of data.
Allow me to point you to a climate scientist (who does not deny global warming). Her site is quite comprehensive. Avoid the comments section.
www.judithcurry.com
What kind of cycles are we talking about? I'm under the notion that typically the only effect that should not be captured should be seasonal change with yearly periods.