First, there’s this, from Michael Tsai’s blog:

I think the best thing that can be said for subscriptions is that they’re honest and mostly align everyone’s incentives properly. Customers will essentially vote with their wallets, on an ongoing basis. Developers who maintain and improve their apps will get recurring revenue. Apple will get more revenue when it steers customers to good apps. Over time, more of the money will flow to the apps that people actually like and use. My guess is that the average customer will end up spending more money on fewer paid apps. Some apps will become more sustainable, but others will be culled.

And this, from Pixel Envy:

Once the uncertainty about which apps are eligible for subscriptions is cleared up, I think there are going to be specific qualities that justify a maintenance-style subscription: quality apps for a particular audience, made by indie developers who issue regular updates. And I think it needs all of those factors. Users aren’t going to pay for an app maintenance subscription if the app is of mediocre quality, or if the developer is a huge company, or if it is infrequently updated — or, at least, does not have some sort of ongoing justification for its subscription cost.

I think Apple should let app developers price as they see fit: a one time fee, in-app purchase (clearly marked), or subscription. Give this new subscription model some time to ripen, use that time to learn. And just reserve the right to reject apps that misuse subscriptions.

By allowing app developers to choose for themselves, they will help flesh out this new scheme and, who knows, an innovative new solution might emerge.