I don't agree that, "proper breeding takes 3-6 years to do it right, otherwise your just a pollen chucker." I disagree with that timeline. A breeding project should have a goal, for most projects that is homogeny. I agree that a typical route in the past has taken years. I also agree that the true agricultural-tek F1 hybrids should take at least 5 years to even be close to being done.
How do you arrive at the conclusion we should only be breeding with landraces, when in theory, landraces do not exist in a vacuum any longer? During the past 20-30 years, I don't even think 0.01% of cannabis breeders truly understood the botanical implications of geno/pheno/chemo types or the definitions of a varietal or cultivar. There are a handful of folks that do apply these types of standards with rigor though - my point being that, these folks, they are the modern breeders, not the folks who just happened to fall into good cuts in years past. Not shaking a stick at them, they were the cutting edge of cultivation too - just not now.
With both the flavenoid topic and drift/etc. correction by a plethora of means becoming more clear every day
If you are breeding formally (on paper, to be audited or reviewed or to be used in litigation) you treat it as a mix between pharmaceuticals and botany. You'd use a Target-Product Profile to define your goal, then Robertson-Price maths to move forward based on your sample size. You can make a defendable argument that stands up to scrutiny (math scrutiny, not opinion scrutiny) even with a sample size of 20 or so...
A wave takes 6-9 months. There are a ton of concurrent steps and feedbacks into the rest of the process but you can turn over this entire process many times in 3-6 years. IMO, the two things that would make it take longer would be stress testing (if thorough) and/or low resources (hands/space/etc.).