Most media productions -- almost all -- are not profitable. They don't attract enough viewers to make back their production costs. A few productions are immensely popular and make back many times their production costs.

It's not possible to predict in advance what media will become popular with consumers (and thus profitable.)

Moreover, popularity/profitability does not correlate with production quality at medium and high quality levels. (That is, bad production quality rarely becomes popular. Medium and high quality may or may not become popular, and becoming popular is uncorrelated with production quality.)

Since a) most productions are unprofitable; b) you can't predict which ones will be popular; and c) as long as you're at "medium quality" or above, adding more quality does not improve your chances of being profitable, the optimal strategy for content at scale is to invest in making as many medium-quality productions as possible. The risk involved in making a few high quality productions is much greater.