People are genetically different. While some people may have higher genetic expression of enzymes in certain areas of the body, others may be much lower in that particular enzymatic expression. So, an example of a scenario which might account for your concern is that perhaps some of those 80-90% of people have lower expression of aromatase or, perhaps, lower expression of estrogen receptors in areas that make their hypothalamus less likely to experience increased estrogenic activity due to the consequential increase in testosterone due to 5AR antagonism.How do you explain the studies showing 80-90% of people showing improvements or atleast maintenance in hair density and count when taking finasteride after 1-2 years, most people ride out the shed and then notice significant improvements after a year on finasteride, not 2-3 weeks
Let's look at the opposite scenario. Let's say that a person has high estrogen receptor expression and high aromatase expression in all the wrong places and little to none in the scalp. Now give them a 5AR antagonist. In the scalp, 5AR is bound and there is a local increase in testosterone and a modest increase in its metabolism to estrogens via aromatase. In other areas of the body (but, most importantly, the hypothalamus), there is a much higher increase in metabolism to estrogens relative to the scalp. So, what does this mean? The modest increase in estrogenic activity in the scalp is much lower than the increase in estrogenic activity in other tissues and the hypothalamus. Because of this, hypothalamic downregulation of steroid synthesis will occur well before you are able to experience significant benefits to hair growth and hypothalamic downregulation of steroid synthesis will result in even lower level of estrogenic activity in the scalp that if that person had never take a 5AR antagonist. And this is why some people may possibly experience shedding that simply just gets worse rather than improving.
This is also the reason why some people experience gynecomastia quite easily while others don't. The ones that experience it naturally likely have high aromatase expression and high estrogen receptor expression in breast tissue, but low estrogen receptor and/or low aromatase expression in the other areas of the body (which leads to lower overall estrogens in serum) and lower expression in the hypothalamus. So, the beast tissue gets tons of estrogenic activity, but the hypothalamus is "happy" with its level of estrogenic activity and doesn't intervene to lower steroid synthesis to reduce estrogenic activity.
In my own body, for example, I likely have quite low aromatase expression and/or estrogen receptor expression in breast tissue because even with a supraphysiological concentration of testosterone injected exogenously, I do not experience estrogenic side effects in breast tissue and even if I did, my hypothalamus has sufficient estrogen receptor expression and aromatase expression that it will cut steroid synthesis when approaching estrogenic activity too high for the satisfaction of the hypothalamic feedback mechanism.
So, the explanation is that everyone has a uniquely different genetic expression of all types of enzymes in various tissues and, because of this, can each react much differently to different types of drugs that target these enzymes.