I'm not a fan of big Pharma, but this is horseshit.
Tricyclics are substantially more dangerous than the newer generation of medications, sure you can OD on any of the psych medications, but the newer medications tend to be more narrowly focused than the old ones. Have you ever looked at the listing of things to avoid when it comes to MAO inhibitors?
A lot of the problem with the newer medications is that since they target smaller parts of the brain, it's less likely that any one medication will work properly, but it also means that it's less likely that it will interact with some other medication. For instance you can't take Prozac or Paxil if you're taking stimulant medication for ADHD because they use the same channels in the liver, IIRC.
Ultimately, this is not likely to be a problem in the near future as brain imaging scans to see what exactly is going on in the brain become more prevalent and there's more formal testing of what the medicine is actually doing. At present there's very little attention paid to how much of the medication actually gets to the site where it's needed. Something as simple as an undiagnosed food allergy can result in little or none of the medication making it to the brain. Which also effects how much seratonin, dopamine and the rest are there for the medications to work with.
TCAs are indeed substantially more dangerous in overdose but they are globally more effective as well. More importantly, most people on TCAs won't overdose on them. MAOIs are entering an era where we have both EMSAM (a patch that avoids the tyramine toxicity issue) and more selective, reversible MAOA inhibitors that have comparable efficacy to MAOIs but much better safety and tolerability. The problem with newer drugs is not a "targeting of smaller parts of the brain". Many newer agents are more selective with regards to receptor and channel targets (pharmacodynamics). It's the primary reason some newer medications have fewer side effects. Unfortunately, many new drugs don't actually have fewer side effects. Instead they have different ones. As for the example you use, there is no blanket contraindication for fluoxetine (Prozac) or paroxetine (Paxil) combined with methylphenidate or amphetamine class stimulants. The primary issue with fluoxetine is it is a potent inhibitor of 2D6 which means other drugs metabolized by that pathway may accumulate to toxic levels when administered with fluoxetine. Paroxetine also has a host of drug-drug interactions but NOT with stimulants. The two antidepressants you cite are selective serotonin reuptake inhibitors they share virtually no pharmacology with stimulants. As for brain imaging people have been sold a bill of goods with regards to utility in neuropsychopharmacology. fMRI, PET, SPECT, etc with a host of pretty false-color images are nice to look at but have provided minimal advancement in diagnostics or therapeutics. The primary exception is structural/flow imaging that reveals a great big tumor or bleed/clot but a decent neurologist could tell you the same thing for a fraction of the cost. We do have better technology but what we actually KNOW about the brain is not progressing by leaps and bounds. We are taking baby steps at best. We dont' KNOW where medicine should be for most neuropsychiatric conditions and even when we do (Parkinson) we are limited in our ability to target specific areas. The problem with Big Pharma is their mission has always been to make money. They aren't evil and they aren't incompetent. The old FDA hoop was to prove a drug wasn't lethal . . . and that didn't happen until the 1930s. It's actually a fairly recent rule that companies had to show effectiveness. The conundrum for Big Pharma is that serendipity has driven a lot of advancements (vaccines, antibiotics, antidepressants, antipsychotics) and we just aren't smart enough to solve problems from scratch. And we are too impatient (myopic) to invest the hundreds of billions of dollars and decades of basic research to learn more about human development, physiology and pathophysiology from head to toe.
Drinks used to be served in smaller containers, and society survived just fine. Restaurants started using larger containers to exploit flaws in human psychology, allowing them to trick customers into buying more than they want or need. This is done to make more money, and to hell with the health of the general public.
Or public health officials have been tricked into thinking it's more important for people to be healthy than to eat satisfying junk food and are exploiting flaws in human psychology regarding the correlation between physical appearance and mental state (we are biased towards believing that attractive people are happier).
That's the problem with the "people are stupid" line of argumentation that's prevalent in the nanny state -- it doesn't really explain why we should prefer moving decision-making from one group of stupid people to another group of stupid people.
Your comment on the nanny state misses the point. You aren't moving the decision from one group of stupid people to another group of stupid people. Most people are IGNORANT of the negative impact of excess sugar-sweetened beverages. It is a reasonable libertarian argument to say people have a right to make bad decisions. It's another issue altogether to claim people have a right to make bad decisions without knowledge of it being a bad decision. Every obese person on the planet KNOWS they are fat because they consume too many calories. But the vast majority don't know the typical sugar-sweetened beverage contains 10-15kcal per ounce, the average American consumes the equivalent of 700 8oz SSB per year and that merely cutting a portion of their SSB consumption would lead to weight loss without a substantial change in lifestyle.
Real Users find the one combination of bizarre input values that shuts down the system for days.