I think it's actually important to note the "number of actors working on optimization". I don't think there are that many people that are interested in trying radical new ideas. Most people are comfortable with a certain type of build or at least a certain pilot and really stick to that. I don't see many experienced players that are willing to really go crazy with the types of builds out there. Most stick to tried and true methods.
I'm not asserting you're incorrect, but I'd like to see some sort of evidence that this is true, rather than take it on faith.
I'd have to pull numbers from last year's store champs, but the impression I took away was that there were a lot of non-meta-standard lists brought to store championships, and very few of them did very well. Assuming my memory isn't faulty (quite the assumption), then it seems that either the only players experimenting with list archetypes are poor pilots (an extraordinary claim, IMO), or that there are few to no diamonds in the rough to be found (because they've all been discovered by the community and turned into meta staples).
Note that "one person with a streak" is an aberration by definition.
Biophysical is obviously a great Defender pilot, but let's keep in mind that his regional matchups had no Brobots (just a lone IG in swiss), Soontirs, Decimators (or EU turrets at all), until the top 4 loss to KO. Maybe he'd have cleaned house against those kinds of lists too, but the issue is that we don't know, and we shouldn't extrapolate from a single data point. That's why I was asking for the number of occurrences to be included.