"Future studies of novel antidepressants with acute psychoactive effects should make stronger efforts to mask treatment assignment to minimize the effects of subject-expectancy bias."
Instead of pondering if their method is at fault for their negative result, the conclusion is to double down and working even harder at masking the effects.
That line just says people should do better at this in future to get more accurate results. It's not a statement about the ultimate form the treatment should take.
I believe benzodiazepines have also been tried as an active placebo but I don't remember what the results were.
No idea about deliriants like atropine/scopolamine, but in low doses, they were traditionally used to spice up beer (hence Pilsen named for Bilsenkraut, black henbane), so I could imagine short antidepressant effect there too.
It does, and this observation led to the development of dedicated antidepressant drugs. There are several such cases where interesting side effects turned out to open a world of possibilities, including an anti-TB drug that also led to antidepressants, and of course Viagra.
"find something that has a similar trip but does nothing" is actually equivalent to the original problem.
But it's also doubling down on the assumption that the subjective effects aren't the source of the benefit.