As a general rule, the parts of your staff that scale O(n) with revenue should be cost sensitive but the parts that scale O(log n) or O(1) should be quality sensitive.
As long as they can keep their growth rate up, the ratio of their staffing expenses to revenue should go down over time, regardless of how high the original denominator was. Thus, if you're a high growth potential company (which, if you compare the size of PLoS to Elsevier, seems plausible), you should pick the city that maximizes your growth potential.
Since traditional publishing was such an intrinsically distributed process in the first place, PLoS has an unusually easy time building a distributed workforce for all of its journal specific workforce like editors which keeps those costs low. At the same time, the team working on functions that span all journals like software or marketing can centralize in a city like SF which bats above it's weight in those areas.
There are highly successful super low-overhead journals that do not advertise, do not pay 6-figure salaries. See http://www.jmlr.org/ for example.
If you receive a 300k$ a year grant, then $1k per paper is 0.3% of your budget. And it is not uncommon for medical researchers to get a lot more than 300k$ a year.
...
Now I want to end on the issue that seemed to upset people the most – which is the salaries of PLOS’s executives. I am immensely proud of the executive team at PLOS – they are talented and dedicated. They make competitive salaries – and we’d have trouble hiring and retaining them if they didn’t.
From the outside, it seems obvious that having a hundred-person office in the one of the most expensive cities in world is simply incompatible with having a low-cost high-quality international journal. While it's understandable how they ended up in this situation, the question now is whether their original mission is important enough to sacrifice the "good thing" they have going.
It seems unlikely to me that the current executive team would ever decide to cut their own salaries in half (or more) and decamp to a lower cost location. But it also seems certain that there are lots of talented people in more affordable locations who would have a good chance of doing an equally excellent job at a much lower cost. But given the (possibly legitimate) self-interest, I don't see any way they can make that transition.
Perhaps instead of trying to cut their costs, their best hope going forward is to keep their prices and profits high, and use the proceeds to fund some new completely independent "startup" journals to achieve their admirable original goal? And perhaps dedicate what they can of their current team to developing software and processes that others can use to run a lower-overhead journal?