That said, I would still say that there are plenty of ways to improve (what I would consider) "comfort" in your work environment; however, they do not need to be reimbursed by the company (moving this back to your original post).
Likewise, as the other reply to this post mentions - cost.
Simply put, what is the cost involved and relative productivity gained/lost from each layout design. What about experimental designs? Without this type of information, it is difficult to convince a decision maker to improve these conditions. Those metrics need to be quantifiable, not anecdotal (I get more work when there aren't conversations around me vs. I resolved 10 bugs in the open office environment and 25 equivalent bugs in a cubicle).
This gets into more the MBA and Information Systems world of quantifying intangibles in tech (like productivity), but to obtain this in an environment that would be deemed acceptable for experimentation, you'd need a corporate culture interested in experimentation - because not every layout will work for every group or every person in said group. Even then, being a group that opts into experimentation is not indicative to a real work environment, and so better results would require a new cohort for each experimental layout.
Regardless, without decision maker buy-in and quantifiable metrics that can beat the cost/productivity metrics of open-office, articles like this only serve to form an echo chamber.