> Second, you can't really cheat your biology in the long run, because your brain will always compensate for any psychoactive drug you consume.
Not true; this is a common pop-sci myth. Nearly all neurological signal processors have numerous types of inputs and outputs. While it is true that the body usually alters drug sensitivity so as to counteract the drug effect, it simultaneously alters many other signal senders and receivers. (You can read about protein kinases and nuclear receptors, which are famously broad spectrum, to learn more about how complicated things are.)
> If you take a drug that increases the amount of dopamine in your brain, then your brain will decrease its density of dopamine receptors to compensate.
Which, of course, does not mean that the drug effect is magically cancelled out, but rather that the cell becomes less sensitive to sudden bursts of dopamine from its neighbors. The drug might also affect not simply the magnitude of receptor sensitivity, but the rate at which the receptor desensitizes during continued stimulation, as well as the firing pattern of the neuron.
Even the receptors themselves are complicated. Most neurotransmitters are received by a family of several distinct receptor types: stimulating one type may make any imaginable change on the other types. Many receptors are made of several interchangeable subunits: the levels of the subunits can change independently, tuning complex patterns of activity rather than simply compensating for the drug. Many receptors are actually receptor complexes that receive several neurotransmitters at the same time: understanding these is presently a scientific tar pit. (IIRC there are several important dopamine-serotonin receptor complexes. The NMDA receptor complex is famously sensitive to a slew of things.)
> And if you increase the ability of your receptors to bind to dopamine, the amount of dopamine in your brain will decrease.
You might find it enlightening to study the difference between postsynaptic receptors and presynaptic autoreceptors.
> Virtually every longterm study on every pharma drug shows that you are worse off in the long run than if you'd never taken it at all, ...
This is not even remotely true. Many drugs are conclusively proven to be very effective (assuming proper patient selection). Popular stand-bys, like those for inflammation and blood pressure, are popular precisely because they Just Work. For as long as you keep taking them.
The central nervous system drugs work too, although the statistics are less impressive. The main problem is that the available drugs are outnumbered by the brain's possible failure modes, so the efficacy rate tends to lag. A related problem is that the drugs tend to be broad spectrum, so it can be tough to dial up to an effective dose without running into unacceptable side effects.