> 1. Why wasn't it successful?
Because the desktop as a target for casual gaming is pretty much dead and the social network desktop web site as a portal to web-based casual games is likewise pretty much dead; Google+ Games was an implementation of Facebook's social-network-web-site-as-portal-for-web-based-casual-games model. A model which even games that integrate with Facebook (for signon, finding opponents, and posting to the Facebook stream) are abandoning in favor of native (usually mobile) games interacting through APIs but not integrating with the social networks Web UI. That model was past its peak when Google+ Games was introduced, and is pretty much dead now (and G+ has much less legacy motivation to retain support for it than Facebook does), so its out the window.
> 4. What are the possible n-th order effects of having a Games ecosystem, even if it isn't seemingly "successful" in isolation?
Irrelevant, because Google isn't abandoning the idea of having a "Games ecosystem", even one with interacts with Google+ deeply, they are just recognizing that the way to do that isn't to restrict that ecosystem to the G+ Web UI as a portal. That's why the announcement that Google+ Games was being phased out was pretty much simultaneous with the announcement of the availability of the cross-platform (iOS/Android/Web) Google Play Game Services API. (Which is integrated with Google+ -- it subsumes Google+ Sign-On.)
> 5. What, if anything, does this say about the possibility of an "apps" (non game) ecosystem within G+?
The apps-within-social-network-web-interface model, which was dominated by games, is basically dead in favor of apps (including games) that interact with social networks via an API but which are not web applications integrated with the social network web API as the exclusive portal through which they are accessed. So, insofar as you refer to that model for G+ when you refer to an apps ecosystem "within" G+, you should probably take the shutdown of G+ games as a recognition that Google has little current interest in that model.
As far as an ecosystem of apps (both web and native [especially mobile, in the latter case]) which integrate with G+ features like the message stream, Hangouts, profiles, etc., the recent pushes behind G+ for SSO, the Moments API, the Hangouts API and the move to raise the profile of Hangouts, etc., should be a clue there.