I wish I knew chemistry and physics, etc. better so I could understand the molecule threads those two do.
Well, Sci Hub, Google Scholar & PubChem combine into a very powerful (free) learning tool.
If you can find and afford an old copy (version 11 seems best) of ChemOffice, that's amazingly powerful. But if not, ChemSketch and the other on-line tools that ACD/Labs are free and are still jolly useful.
I suppose if money is no object, a Reaxys licence that includes unlimited searches is the MOST powerful tool but is costly. I specified older versions of ChemOffice because I'm told the newer versions link you to Elsevier and other publisher so yes, you can read the paper almost instantly, but unlike using Sci Hub, it bills you! Few things are more upsetting than spending anywhere from $30-$80 for a paper only to reaslize that only one sentence is of value.
I believe there are also some more advanced tools but they really are designed to extract the maximum amount of money out of researchers.
As a class, I suggest that the phenethylamines represent the easiest class of psychoactive to explore. Isomer Design even has a search engine in which you provide the basic scaffold and it displays every compound in it's database that contains said scaffold.
But I would estimate that as with everything, it's 1% inspiration, 99% perspiration. It's so easy to just draw a chemical and assign a name and theoretical activity. Part of that 99% is working out a facile synthesis.
I don't feel it's hard, but it takes a long time and get ready for doubt. Because as Karl Popper noted, an experiment can only disprove a theory.
If all else fails, ask. But do the 'all else' first as being asked something that a Google search will find in ten seconds can be frustrating. But we are all learning. Nobody can legitimately claim to know more than a fraction of all of one science, let alone all of the sciences that intersect with chemistry. So there really are vistas of information that has yet to be extracted from data.