A short while back, maybe last year sometime, I had a brief exchange on a blog somewhere relating to the development of dynamically placed type for journalism maps at places like NYT (in fact I think I was talking to one of the map designers for NYT about this). I don't recall the details but I know they were innovating an approach to placing text on static maps programmatically, as opposed to manually, despite the map being static. Seemed odd to me to code labels placement for a non dynamic map, but I'm an old school mapper. There was some time savings for them.
As I'm looking around at map graphics today for a new Cart in GIS lecture I've noticed that NYT maps in particular exhibit a strange detail when you try to view them separately from their web pages (by R clicking and going to View Image...), the background static maps opens up but the seemingly static labels disappear, making the stand alone map useless in many cases.
Compare the two screen shots form a NYT story done a few years back featuring an ISIS territory map.
NYT maps are usually carefully sized and rendered in their stories that calling up the map separately is not usually needed, although I tend to do this regardless. But across the web many online static maps are often too small for close reading and by opening the image alone you can often get a much better view of the maps details. Dynamically placed text in these cases seems to be a threat to that "feature". Not sure its a very important issue, but something I noticed and was curious how folks here perceive this.