making attribution claims easier
Basically, the advantage is that attribution as required by GFDL and CC-by licenses becomes more or less automatic assuming the identity or sign-on is not cracked: people login once under a persistent pseudonym or perhaps their body name if they voluntarily disclose that, and thereafter, whatever is done from that login is properly attributed.
This has many advantages, few drawbacks, and tends to help overcome the problems with strict IP number based accountability and with identity disputes arising from the many different user name spaces. It should also solve the GFDL corpus problems with the attribution required by the GFDL itself and reduce the impact of Wikimedia corruption on the GFDL corpus.
By reducing assumptions about alleged and collective identity, i.e. the allegation or assumption that User:X on Wiki:X "is also" User:X on Wiki Y, it may reduce identity disputes. For those who used such a voluntary common ID or single sign-on it would be slightly more difficult to impersonate them.
Ensuring all wiki user pages alleged to, or claimed to, belong to "the same person", and keeping these claims or allegations correctly attributed, is further simplified by reliance on a standard wiki URI for all such pages. Typically the Language:User: space is reserved for this on MediaWiki-based services.
factions still required
However, for clashing allegations of identity, with deliberate deception, only a faction system can possibly work in the long run. See en:Wikinfo:faction for a proposal for this that would suit pan-GFDL editing, anden: OurAnswer:faction for an experiment in this method.
(insert example of the variety and complexity of problems arising from lack of such a standard, exploitation possibilities caused by identity confusion)
Such a standard would not resolve other collective and alleged identity questions, where one is dealing not with self-alleged identity but with other-alleged identity. There is a fairly complex interaction between the question of factions and that of power structures, including the role of sysops and other such empowered users, and those they specifically disempower, called "trolls" and "vandals". Much of the complexity is re: trolling:
Avoiding the issue of an interwiki identity standard will almost certainly lead to various assertions that "X is a troll" where X is some inexactly defined cluster of concepts or styles or IP numbers, leading to what are called echo chamber allegations floating all over, and even to other large public wikis. Accordingly this is probably an important issue to address in GetWiki, which ideally would support jabber.org logins.
Wikinfo could have led, but won't
This is a huge problem on Wikipedia and somewhat on Wikinfo, though this seems less likely, since sympathetic point of view requires fewer intra-article clashes. Of all GFDL corpus access providers, Wikinfo may be the least likely to actually have this kind of problem. If GetWiki also monitors the whole GFDL text corpus at the standard wiki URIs, and the wikitext standard supports the RecentChanges and DeletedPages log's conventions, it would be easy to have some common address other than simple IP numbers across many large public wikis. However, a proposal for GetWiki recognition of one ID, preferably jabber.org ID, for all of GetWiki space, was censored by the particularly stupid GetWiki developers, who seem clearly only interested in their own commercial benefit.