In short the early adopters will do well in the near term. Then we'll all be worse off.
Most accurate quote here. Short-term gains will turn into long-term losses.
A la carte sounds great to a lot of people because they're assuming that they'll have today's full suite of channel and program options to choose from. However, in the long term, many (most?) of those channel options can't be sustained under a direct subscriber model, and the ones that survive will only be the very largest broad audience channels (i.e., the circa-1985 cable lineup of ESPN, MTV, CNN, USA, TBS, etc.). That defeats the entire allure of a la carte in the first place. Is it truly a la carte if you just have 20 channels to choose from at the end? That produces less programming for Netflix and Amazon Prime (which, even though they're producing some shows on their own, are still largely dependent on third party IP), which reduces the choices on those platforms.
Now, one could argue that if channels can't survive by the direct subscriber model, then they shouldn't have existed in the first place. I understand that from a detached economic perspective if we're talking about the "widget industry", but note that most people don't consume media in the way that they order dinner (or some of the other examples that I've seen) at all. Ordering TV channels a la carte isn't the same as ordering a cheeseburger for one meal. Instead, it's akin to ordering a cheeseburger for every meal every day without variation and knowing that is ALL that you want. Most of us don't consume media that way - we'll nibble on a drama here, a comedy there, a sports event one evening, a news program another evening. That would apply to TV, surfing the Internet or even on-demand choices like Netflix.
Note that Netflix streaming isn't attractive as a pay-per-view service (which is true a la carte) - the whole value proposition is that it has a critical mass of an entire array of programming to offer lots of different types of people. This isn't that different than basic cable - it just comes in a different package. The costs of producing House of Cards and Orange is a New Black are getting subsidized by the subscribers that mainly use it for kids programming (which, if you have young kids, is probably where Netflix gets the most usage) and movies just as you see ESPN subsidize the costs of other Disney channels on basic cable.
So, cord cutters might be getting some programming cost arbitrage right now because Netflix streaming (which, to be sure, is a GREAT service for the current price) is still in growth mode making little to no profit, meaning those cord cutters are getting a lot of programming paid for by the basic cable model at a less than basic cable price for the short-term. Simple economics says that this can't last forever - either Netflix prices have to rise in order to continue to maintain the same level of programming (and shareholders will inevitably want to see profits instead of subscriber growth) or Netflix will need to reduce their programming options (which reduces its value as a basic cable substitute).
Now, this might inevitably be the model that we'll all have to live with. I just don't see it through rose-colored glasses through the long-term. A la carte choices are only attractive to the extent that such choices actually survive and exist (or else, all we'll end up doing is paying a la carte for the circa-1985 cable channel lineup in the long-term). We might pay less money for less programming. We might pay more money for the same programming that we have now. In no way, though, will we pay less money for the same programming that we have now under the a la carte model. Too many people are being tricked into thinking that the latter is somehow possible in the long-term.