Conjecture, hypothesis, theory. I've run countless experiments on myself. Each one usually took a year to see if there was a valid result. We can't know what how our bodies might respond to inputs without running an experiment. My wife and I joined our first gym in '79 so we could go salmon fishing in Bristol Bay. I've run many different programs on both of us. It's complicated. In general, nothing interesting happens in 5 weeks. Although exercising will always give a better result than not exercising, it's a particular result that one is seeking. I'll start a program in October, might see a result or lack of same (still a result) in April. When one is already fit, it's not a question of getting fit, but rather what sort of improvement might be seen in a particular portion of a sport. A lot of it comes down to training ganglia in a particular skill. Takes time.

One sees folks who might try to increase their accustomed cadence from say 80 to 90 and find that they're worse at 90. Yep, might take months to become accustomed to the new cadence and actually see if it's better or worse. Of course by then, one's general fitness might have become better or worse, clouding the issue. It's not 5 weeks simple. As you say, 5 weeks is common, just not definitive, that's all. IME studies which don't last for months can probably be disregarded.

Looking around, I ran across this website which approximately agrees with my experience: https://roadcyclingacademy.com/weigh...-for-cyclists/
I used to take an hour full-effort spin class, then hit the weight room immediately after. Oddly, my results in the weight room on those days was the same as when I didn't have the spin class. I was in my early 60s and quite fit.