first of all
Analising small samples is possible and not to be considered a `terrible argument` as you stated before
Depending on the disease you will have minimum number of patients needed to research... prostate cancer, the most common cancer on male, you will need like between 1.000 - 5.000 patients to make an impactfull research... wegener`s granulomatosis (rare kind of vasculites), on the other hand, you will need way less patients and still be significant and important for the medical community... small samples wont hurt its validity... you can validity 2 studies, one with 10.000 patients, other with 250 for different diseases
second of all
majors are such important games, it dictates the meta, teams saves strats to this tournament, whole community of CS-GO stops to watch it... how cant you consider it the most important stuff to analise???
topson is considered bigger in dota history compared to arteezy... look how small the sample of topson is compared to a decade of arteezy`s participation in The Internationals...
third of all
top 5 hltv actually displays from lots of variables on team forms and achievs, they are enhancing the list better and better with time... you also cant use that point of view against me if you are using the same top list...