THE OLD ORDER CHANGETH ...

The old order changeth,
yielding place to new.
—Tennyson, Idylls of the King

Quality content vs. fluff may be one of the ostentatious reasons for the fissures in the community, though not the sole one.  There have always been serious and non-serious writings on Gather. Often frivolous writings got disproportionate attention while some gems of articles languished in anonymity.
 
There have also been elitist vs. non-elitist battles; not in the form of class struggles per se, but more as abrasive reactions to arrogant postures.

The old order was typically of a feudal nature.  The arrival of Carl Andre upgrade and the mass induction of new members eroded the small town prominence of feudal lords and ladies. The 'points' bait, to which members had by then got fairly hooked, went askew with the misuses, defiance and disorder attendant of a burgeoning membership.  The captive readership scattered helter-skelter because of temporary disorientation caused by the new interface.  The new voices, already exposed to other blogging sites, asserted themselves aggressively.

The proliferation of 'non-articles' is, indeed, the problem of Gather management who have to provide the storage space for all the stuff.  The expansion of membership base, the financial viability and the sustainability matrix are also management concerns.  Members can further the cause of building a better Gather by fostering goodwill rather than infighting.  That will ensure the retention of the pleasant diversity without endangering the viability of the platform.  The quality concern can be addressed in an altogether different and ingenious manner.

Tom Gerace, while introducing third-party social book-marking tools, promised that a system upgrade that will include Gather's own tools with the ability to share contents via a technology called RSS is on the way.

While the universe, as a rule, favors mediocrity, a shareable book-marking tool can act as a bubble-sort algorithm for separating the nuggets of quality out of the mountains of mediocrity.

It will be an innocuous feature that will simply facilitate members to organize and bookmark the articles of some lasting value as per individual predilections.  It is the shareable feature of the facility that will work as an unintended 'peer recommendation'.  People who have less time and patience to dig into the archival heap for articles of quality and value will try to draw into the resources of known connoisseurs whose taste and judgment they trust and value.  Such a system of peer recommendation will enable individual members to build their own libraries.  The community interaction will increase. Though the chat-and-game articles might still make temporary splashes and steal the limelight as the 'Highest Rated' and 'Most Discussed' articles, the long term churning effect of peer recommendation will ensure the discovery and survival of writings of enduring quality.