Is every flouting of orthographic norms an instance of neography? It seems most scholars in the field would say yes, but I think there’s an important distinction to make: that between intentional and accidental neography.
When we discuss neography, we’re usually talking about intentional neography. This is when we might write <thru>, <tomorrow>, or <krazy>; we’re indicating informality or trying to establish something about our identity or communicative goals through our spelling choices.
But what about when we write something like <informatoin>, accidentally transposing letters; <caling>, forgetting a letter; or <oractice>, missing an intended letter and hitting a different one? These can certainly be considered neographic, given that they diverge from the orthographic norm—and they do, after all, signal something about the exchange: that there’s a time constraint, for example, or some other external pressure.
Although accidental neography can seem little interesting at first, such expressions merit study for a number of reasons. For example, they can show a speaker’s underlying pronunciation of a word, which would be otherwise obscured by the orthography. If a hurried person writes <cot> instead of <caught>, for example, we might speculate that they’re subject to the Northern Cities Vowel Shift. Or if we come across <segway> instead of <segue>, we can get a less equivocal picture of how that word is pronounced (brand name notwithstanding). If a Spanish speaker were to write <actris> (which should, orthographically speaking, be <actriz>), we can conclude that this speaker doesn’t phonologically distinguish between [s] and [z], which is typical of certain dialects.
Of course, we should note that the boundary between intentional and accidental neography is diffuse and difficult to define. For example, <teh> might signal that the person was typing fast, but it could also be intentional—going the way of <pwn>, for example.