Authentication technology integration for Social API

In DrupalCon Prague social BoF the guys behind fluxkraft have introduced us to their module suite and distribution for automating social media workflows.
The core of fluxkraft is built on top of Guzzle a PHP HTTP client & framework for building RESTful web service clients that is now a part of Drupal 8 core.

The stand out features that impressed us were:

• The module suite and its use and integration with rules was very impressive
• it was intuitive to use providing a great interface based on rules
• Specific implementations to social services (e.g. facebook and twitter) implemented as individual modules
  see https://drupal.org/project/fluxservice

Looking at how such modules and suites of modules can help to deliver the framework for the Drupal Social API / Initiative we pose the following questions:

1. A Pure Drupal solution: Should the development of specific social services integration be handled by Drupal or are there other open source / non Drupal specific projects we should be using that may present a more generic [more widely adopted and supported?] and less Drupal centric/specific approach?
2. Architecture: Is the architecture of fluxkraft the "correct one" i.e is right for all use cases?
3. Standards or module suite: Should we attempting to envisage a component standards based architecture that doesn’t aim to be an all encompassing ‘framework’ but rather provides the ‘glue’ for modular components each providing specific and non-overlapping functionality?

A Pure Drupal solution

Since Munich we have also been made aware of HybridAuth. This is an open source project with a goal to act as an abstract api between your application and various social apis and identity providers such as Facebook, Google & Twitter.
This also has a Drupal module see https://drupal.org/project/hybridauth
There are others out there that try to do the same either as similar projects or commercial services.
So should we be trying to leverage HybridAuth or other solutions?

Architecture

We haven’t spent enough time looking at fluxkraft in detail to do it justice, so the questions are in no way a criticism.
It is worth looking at the stack of the solution and asking a few questions like what are the abstraction layers?
Does the API have a low level integration level that's not rule bound and a higher one for rules implementation?
If the core API was only built on top of guzzle and had good abstraction layers it may be more suitable to one day end up in core.
Is Hybridauth also attempting to be too all encompassing
Should our aim to be to work with projects such as Hybridauth / fluxcraft to evolve standards and more modular approaches and aim to select elements of both.

Standards or module suite

Since the aim of the Social Initiative isn’t to develop yet another ‘module’ but rather attempt to gather together the knowledge and participation of key players in the development of such modules, should we rather be looking at providing the standards and collaborative framework for the wider integration of such modular solutions?