I’ve noticed a strange tic among academic writers, a tic to which I fall victim myself: a compulsive overuse of connective particles. In some cases, it seems that there must be some type of connective particle in literally every sentence of every paragraph. It’s as though we can’t trust our readers to infer that our sentences are related to each other by virtue of being placed in sequential order in the same paragraph, etc.
As Voyou said when I voiced this concern on Twitter, “what makes it clear that it is a tic is that they’re always the most general connective particles (like ‘thus’), so they gesture at a connection without saying what the connection is.”
My hypothesis as to the origin of this tic is the increasingly frequent experience of over-hasty and dismissive readers of academic writing. Reading half-attentively, ready to jump on any pretext for criticism, these readers simply cannot be trusted to, you know, read our work. Instead, they dip in and out, reading every third sentence while scanning over the rest.
And this is where the connective particles come in. They remind the inattentive reader that each individual sentence can’t be taken in isolation and criticized as though nothing else had been said. No, they’re connected to the rest of the piece, which must actually be read and assessed as a whole rather than as a collection of isolated monadic sentences suitable for target practice.
In other words, the “thus” tic is a kind of verbal cringe, a wincing attempt to dodge the inevitable arbitrary criticism. Every time we throw in a “thus” or “in other words” or “that is to say,” we’re saying: “No, see, I wrote these other sentences, too, and what I’m saying here depends on what I said there. You can’t just take this one sentence in isolation!” It’s an attempt to shame the impatient reader into at least glancing at the previous sentence before pouncing on the one their judgmental gaze happened to fall upon.
An und für sich 
This trend may actually be part of “running academia like a business.” From my experience with writing in a business setting, any time a client criticizes something — even if they’re demonstratively wrong — you have to respond. They won’t accept that it’s their fault for misreading or reading overhastily. If they misunderstand or find it lacking in any way, it is definitionally the writer’s fault, not the client’s. Hence you wind up writing in a way that anticipates dumb questions.
The particle tic is probably also produced in part through students attempting to fluff word counts. Another possibility could be the anxiety of sitting in front of a computer and trying to force oneself to write. One finishes a sentence and then types a connect particle while trying to think of what to write next–almost like the typing equivalent of ‘um.’
I wonder if the 18th-century German style of 10-clause long sentences had the same purpose? When eg. Kant makes a claim, more often than not the statement thereof is embedded in three layers of ‘weils’ and ‘alsos’.
Is there any reason to think that 18th-century readers were especially likely to be distracted? This was the era when philosophical journals were introduced so maybe with the expansion of the readership beyond a small circle of scholars, writers at the time didn’t feel like they could trust their readers to follow the connection?
Meanwhile, Augustine can afford 20-page digressions because his readers are all literally locked into a room by themselves for 18 hours a day.
This assessment seems basically right, but isn’t just the case that, as a rule, academic writing deals with difficult topics–some of which require more instances of “thus” and “that is to say” for clarity’s sake?
But to add to your point, there’s also a natural impatience with “secondary” literature (i.e. us) just because there’s so damn much of it. We bear a heavy burden of proof for someone’s (already limited) time in unprecedented ways.
I wonder if the very common deployment, in passing, of “then”–when no previous logical operation merits its use–is part of this syndrome?