Why?
Well, as a clinician let me address each one.
1) The mind is not the brain. That may sound odd, but it is not. We simply do not know yet how the brain and mind are fully related. We can speculate that because serotonin is involved in mood regulation that if someone is depressed, they lack enough serotonin. So we create a pill that in one of several ways regulates serotonin production or uptake. However, there is no empirical evidence that it works. We can not yet measure serotonin levels via blood draws or via spinal fluid or any other objective measure. It is still a 'best possible guess' based on self-reporting from those taking the drug that it has helped their mood regulation. And then there are the unforeseen side effects, discontinuance syndromes, and addictions to such drugs like the benzo class of anti-anxiety medications. Yet, I can also teach CBT techniques to someone, and they will also report a change in mood regulation over time as well. So do we need pills or techniques alone? Both? Why? Why not? Scientists, psychologist, psychiatrists, philosophers, and computer scientists still can not agree on a single definition of consciousness, mind, intelligence (artificial or natural), etc.
2) This leads to the second desire you state. If we are still in relative infancy with regards to the brain/mind question, then how can diagnosis and treatment be stable? Psychology is really only about 100 years old as a discipline. Why did certain treatment methods apparently work in the past but no longer now? What changed? Was it cultural? Was it biological? Why was homosexuality for instance considered a mental disorder up until the 1970's and is not today? If stability had been in play in that instance, then psychologists would not have revised their ideas of what mental health and mental disorders entail. The civil rights issue of marriage would have been a non-issue and not possible today because 'science' would have said, homosexuality is not 'normal' therefore not acceptable. Change is inevitable and that degree of desired stability will be the same balancing act it has been since psychology started. We need a certain amount of stability, and we need a certain amount of freedom to learn more, adjust our findings, and come to new conclusions.
3) Again this ties back into desire number one. Why does both an SSRI and CBT both produce results for an individual with a mood disorder? Which should be the preferred method? Psychiatrists dispense drugs so they will say an SSRI. Psychologists use non-drug methods so that will say CBT. Is one more right than the other? Should patients only have the choice of one method to treat said mood disorder? Even within each of these choices, there are further options. Not all psychiatrist use certain classes of psychotropic medications. Not all psychologists use the same exact formula of CBT - some use ACT, some use MBCT, and others use Dialectical Behavior Therapy. Which of these in either category is the 'right' method. Diagnoses change. Treatments will change as well. Techniques used to 'convert' homosexuals to heterosexuals were appropriate pre-1970. Now we look at them with shame and revulsion as being not just clinically wrong but ethically as well.
We are a long way from any of the goals you express a desire in, and bluntly I just don't think they are achievable. I wouldn't want them to be. One of the big issues that the western model of medicine has is the one-size fits all diagnosis. A diabetic is a diabetic is a diabetic. But only when a single magical number is crossed. Yet, what about the individual whose blood sugar is consistently one point off from the diagnosis? What of the long term affects on their health of elevated blood glucose? Ok, now these three patients have diabetes. Is there a one size fits all treatment plan? Of course not, because each patient is biochemically an individual. One may respond to diet and exercise and drop back into the 'normal' blood sugar range. Another may require insulin shots. The third may do better on metformin.