Equally obnoxious is that these surveys are blind towards receptor subtypes, all of which do very different things in the brain, as well as its bias towards amine neurotransmitters, while there are other neurotransmitters that appear to have more selective pathways.
Finally, this type of hypothesis doesn't explain the "lag" between when an antidepressant raises neurotransmitters and when it achieves therapeutic value. There are studies indicating, for instance, that patients' (lab measure) serotonin levels had nothing to do with the therapeutic value of SSRI's.
My orthomolecular psychiatrist had me get labs for catecholamine levels in the serum, and I really told him that I won't take stock of those numbers until they check my cerebrospinal fluid; besides, the labs are designed to check for excesses only, not deficiencies. They were designed to screen for pheochomocytoma (adrenal tumor). There really isn't enough science to these surveys, and I resent the notion that you can attribute symptoms like these to simply a "deficiency". Notably, why are extrapyramidal symptoms and Parkinsonian traits not on the list? What about elevated prolactin? Curiously, I have myoclonic jerks in my sleep, I have most of the psychiatric symptoms above (but not the physical ones, even though by all rights I should be anemic), and I also have elevated prolactin.
There are better ways of detecting which neurotransmitter receptor ligands/potentiators will have the most therapeutic value, and this is by the empirical evidence regarding which population subtypes respond to such agents.
I'll present the matter somewhat more articulately in a scholarly book I got called "Clinical Advances in MAOI Therapy"