Here’s a great article on generic drugs and bioequivalence: Generic Drugs: Are They Equivalent
The general upshot is, mostly, no, they are not inferior. There are a few inaccuracies going around, so taken one by one:
“as there is really no way to mark whether they are working, other than by the patient’s own descriptions”
This isn’t true for the testing of generics. This is true in terms of a patient’s response, if they say it isn’t working it probably isn’t working, but not in how generics are found to be viable. They are tested via the amount of active drug circulating at various time points after ingestion (bioavailability) , and thus is not dependent on how the patient feels about it.
“If the doctor specifies the brand name, the insurance usually has to cover it.”
Not really about generics, but this is very much not true (at least not in the US). Insurances have their own lists of drugs they carry, and can exclude brand names. And usually they will just not cover all of a certain kind of drug, and generics are suggested because they are affordable despite not be covered, not because they are covered.
And this: “I don’t remember exactly what he said, but there are a lot of technical things that can affect the effectiveness of a drug. I think there are binders, for example, than can block the effectiveness of the drug. There are time release pills that work, and the generic isn’t time release and that messes you up badly.”
Both of those are not true, because if you block absorption, you block bioavailability, which is how equivalence is tested. So you can’t have a time-release pill and a non-time-release pill be bioequivalent. Unless the time-release aspect is false or ineffective, I guess. However, there can be issues with things like allergies, as binders that do not affect the bioavailability can be different.
However, the 20% figure is correct, and that’s because that’s generally taken as the line of ‘clinically significant’. That is, that’s the point at which you’d normally start to notice differences in groups of patients. However, that doesn’t mean that a single patient may not notice a difference. On the other hand, however, most systematic reviews have failed to find a systemic difference in outcomes between generics and name-brand (links in the article I linked), which supports the clinical insignificance of the differences.
As also has been noted, placebo effect can play a role, and psychiatric disorders are many times highly affected by it, more then other disorders because of how it works. As a result, it’s not hard to believe that your friend could be suffering from just a lower dose, just placebo effect, or a combination of the two.