Someone else might choose the word “frictionless.” But friction is not to be avoided or disavowed. “Grace” here is imagined as a secular version of both grace, in that other sense, and frictionlessness.
What is the most grace-ful process of publication, in that sense, under the disposition to which I am subject (including both injunctions and privilege)? Not the “fastest” or that with the least “friction”: it might well be faster and more frictionless to tape a piece of paper to one’s front door or to a lamppost.
GitHub is there for those who have time and inclination for VCS/SCM learning curves; a text editor using the GitHub API can make a GitHub Gist from the current buffer or file, and in a second or two of network latency, if you are privileged to enjoy the access for which that latency is a signature, you have created a document with a public URL, displayed with a Markdown (etc.) parser, including source text and version data. Variations on that URL via, e.g., gist.io (no longer working) or roughdraft.io or bl.ocks.org provide other presentation formats. And the gist is a repository that can be cloned and committed to and pushed and so on.
At the other end or extreme of the process or continuum, Graceful Publication might also mean not (just) giving in to the mutilation of one’s text so much as making a point of personally publishing both offcuts and preprints, with as durable URLs as this world allows (admittedly, no more than it allows to anything else), and using semantic versioning to preserve public reference to one’s preferred version as author: “beta” can be used for the version submitted for editorial manipulation, with 1.0.0 being the preprint — the revision of the beta that you are comfortable releasing, because it preserves your intentions — and 1.0.0-rc[1, 2…] made permanent branches with revisions made only because you have to, ending with a “release candidate” that becomes the postprint, cherry-picking only the ones you truly value for subsequent versions (1.1.0…) of the preprint.
I agree with anyone who says that using something like GitHub to publish a single-authored essay consisting of a few small files is like using a [large and sophisticated system designed for a massive and complex job] to [perform a very small and simple job]. But there is no denying that a public source text repository exposing a version history and providing “builds,” perhaps automatically through continuous integration, makes what I describe above fairly grace-ful to do.
Some have already considered the latter:
There are more ambitious (perhaps) ways to consider all of this, though they are even more problematically mixed up with the cyberlibertarianism of Silicon Valley corporate culture. The most I’d suggest is cherry-picking the many good ideas here: