Author here. I don't disagree with you, especially in the second paragraph. To some degree I think it was smart for Netflix to make a big splash this early on. But the perception that they're "paying for quality" is a bit suspect, given that Hollywood itself -- ostensibly "paying for quality" in its entire process -- has a 98% failure rate on developing projects. Quality talent doesn't always yield a quality product in the creative business. Even the best writers, directors, producers, and actors have their failures. I have no reason to suspect Netflix was any smarter in its approach to developing
House of Cards than networks would have been -- though I could be wrong (and hope I am).
At the end of the day, the $100M for HoC was a big bet. It was a signaling tactic, a marketing tactic, and a bet on top-tier talent. But it was a bet, nonetheless, and an expensive one. That it seems to have paid off (to an extent? jury is still out) shouldn't fool us into 20/20 hindsight bias. It could just easily have busted.
"So as I understand it, the author is suggesting that NetFlix produce a bunch of low-rent pilots (or half seasons or so) throw them against their wall of subscribers with a bit of recommendation-targeted magic, see what sticks, then spend the money to produce the "grown up" versions."
Not necessarily, and I should clarify here. I think pilots can be shot for a lot less than they are, and I think audiences are proving themselves quite receptive to previews and other short-form "pilot" systems as proofs of concept. A lot of movie trailers are increasingly being made before the movies even start shooting ("teaser" trailers). I think there are ways Netflix can take the guesswork out of the development process, even if production is still an expensive endeavor.
You have to pay for a quality product, but the system right now also pays heavily for the risk in the pre-production process. That process can probably be de-risked considerably.
I would not recommend testing 100 half-baked concepts to see what sticks. But I do think Netflix could test or tease concepts in development to determine which ones to develop, and/or how to present new shows to the right early adopters.
Something I probably wasn't as clear about as I could have been: half the battle in finding a hit is developing an audience, not just a show. Netflix has better audience-development capabilities at all stage of the production equation. It doesn't need to guess quite as much as networks do.
Maybe -- to your point -- Netflix's audience-segmenting capabilities give it the ability to play the big bets game more than networks can. But historically, people have been shockingly bad at predicting what will actually hit and what will miss. Star Wars was laughed out of every major movie studio in town, for instance, and at the script level, it sure didn't look like "quality" to the eyes of the execs reading it. On the other end of the spectrum, a show like Firefly -- which fans (including myself) consider one of the most brilliant things ever made, has repeatedly failed to gain mass appeal beyond its limited niche, despite many attempts at re-marketing home video, a feature film, etc. The operative lesson here is that "quality," as an objective concept, is hard to parse and hard to pin down. "Quality" is more of a function with a lot of variables, and many of the biggest variables are on the production side and the audience side -- not the development side.
Tldr: People are over-investing in development, when they could be investing more heavily in production -- whether at greater scale, or in greater depth per project.