For example, how much do the "95%" of walled garden, mobile app developers make.
The discovery process for such apps is also interesting. Does the arrangment of apps into "Top 20", "Most popular", etc., type lists, as seen in Apple's AppStore, boost the success of the "5%" at the expense of the "95%" that become less likely to be discovered.
There is also an analogy to SERPs. Users are encouraged to consult only the "top result" on page one. This structure plays well with an ads sales business, where advertisers compete for a spot on the top of page one. Search results not in the top 5 are like the "95%", unlikely (and not intended) to be discovered.
To me, these "businesses" resemble a lottery in some respects. Winners are promoted but "99.9%" of players are not, and will never, be winners.
Granted, this is a subjective viewpoint I have had for a long time, and the necessary evidence to prove/disprove it is not made public, but the Twitch discovery seems to support the idea.