This is exactly what pushed me away from pursuing cosmology. One of the most important 'discoveries' in cosmology was that of cosmic inflation [1]. The issue is that if you model the big bang, what it creates is nothing like what we see. Regions of space that wouldn't have had time to have become causally connected (nothing could have gotten between them even traveling at the speed of light since the time of the big bang) are causally connected.

So the solution? There was some very specific period of acceleration, faster than light acceleration to some very specific degree, at some specific point right after the big bang which then ended at some other very specific point, at which point everything returns back to normal. What's the logic for it? Well if we do this, then what we see matches what we expect to see. It's an absolute retrofitting of numbers to make a model fit reality. And not only does that constitute good science in cosmology, it was sufficient to win a Nobel Prize.

It is literally unexplained magic being taken as a fundamental concept of science. And the worst part of it all is that it's completely unfalsifiable. We only have one universe to observe. So if it turns out that the discrepancy is actually being caused by many smaller effects we stand near 0 chance of ever discovering because each one of those small effects will have a million flaws, while the perfectly retrofitted inflation theory has "none." At worst they'll even get absorbed into the inflation theory model magic. Just add more epicycles!

[1] - https://en.wikipedia.org/wiki/Inflation_(cosmology)