Do you have a particular example of this?
https://www.thelancet.com/journals/lanchi/article/PIIS2352-4...
Published August. Based on the Covasim model, paper describing it here:
https://www.medrxiv.org/content/10.1101/2020.05.10.20097469v...
The model takes its IFR data from Verity et al (see table 2), which is about 4x too high because it's taken from Wuhan evacuee data in January. The Verity paper that calculated this had access to a bigger and better dataset, the Diamond Princess cruise ship which they acknowledged they had looked at, and which yielded a lower IFR, but they didn't use it:
https://www.nicholaslewis.org/covid-19-updated-data-implies-...
"The Verity et al. CFR estimates were derived primarily from Chinese data, which reflected non-random testing ... When Verity et al. was prepared, the final death toll was not known. The data available only ran to 5 March 2020, at which point 7 passengers had died."
There's an analysis of the school re-opening paper that explains all this here. It also claims the paper used values for k (over-dispersion) from February, when there were hardly any cases on which to calculate that: https://lockdownsceptics.org/schools-paper/
Here's another paper that used an IFR of 1% in May:
https://www.nicholaslewis.org/did-lockdowns-really-save-3-mi...
And another paper that assumed a 1% IFR in June:
https://github.com/ptti/ptti/blob/master/docs/PTTI-Covid-19-...
Analysis: https://lockdownsceptics.org/new-ucl-paper-on-contact-tracin...
That one didn't even bother to give a citation for its source, just assumed a value that was 4x too high.
Here are some current, reasonably up to date and sourced IFR estimating papers: https://swprs.org/studies-on-covid-19-lethality/
Epidemiologists know IFR rates fall over time because they always do. Yet they continued to use the earliest calculated value they could find, based on a tiny dataset that wasn't even the best available, because it gives the highest value and that lets them make the most dramatic results. This is a systematically untrustworthy field, they literally do not care about their own reliability at all.
Epidemiology has so many massive cultural and methodological issues you could write an entire book about it, maybe one day I will. It is absolutely de rigueur in this field to make predictions without ever bothering to go back after an epidemic played out to study how well those predictions matched reality. They cherry pick data at an absurd rate: if they can get a more dramatic paper by using data that's 8 months old and based on a sample size of 7 people when they could use data published last week and which has a sample size of hundreds of thousands, they'll happily pick the former every time. Nobody in that field will notice or care, and they'll throw in a misleading citation or two to disguise what they're doing. There is no code quality control: these people happily publish papers based on models filled with memory corruption errors that can't even replicate their own output.
As for ideology, the fact that you can't explain what climatology sceptics believe should give you pause for thought. You appear to be projecting: a blind devout belief in "scientists" (they aren't really) is itself an ideology. Rather than engaging with the concerns of people who are pointing out problems in the published literature, you're just blowing them off as irrational.