David has a hard, high-80s changeup, usually about 4-5 mph off of his fastball. He has thrown only two this year (both to lefties). Both pitches have been balls - one was about a foot outside and the other bounced up to home-plate. Last year, Robertson's changeup was in the strikezone less than a third of the time, and of those pitches out of the zone, batters swung only once. So, it's understandable that Robertson would want to tighten up his changeup command, especially given that he only has two pitches that he can rely on (fastball/curveball).
Chamberlain's changeup is pretty slow, averaging about 10-11 mph off of the fastball. With his move to the bullpen, Chamberlain has been fastball/slider/occasional curveball; he has thrown just one changeup all season long, and none since the first game of the year. Last year, he used it 4.4% of the time (120 pitches). Though he got some encouraging results on it last year (.315 whiff rate on swings, mind the limited sample), it doesn't look like he and Jorge Posada/Francisco Cervelli are willing to go with it that much.
Apparently, Hughes' changeup development was earned him the final spot over Chamberlain in the Yankees' 2010 rotation. It gets similar action to Chamberlain's, though these days he throws it a bit harder - 83.3 mph average as a starter in 2009, 84.7 mph average this year. This might be due to the new grip he's using this year. Despite the pitch's apparent improvement, Phil doesn't show it that much. He's only thrown 26 all year (1.9% of his repertoire), and has only thrown it for a strike (called strike, swinging strike, or hit into play) 9 times. Hughes is having an impressive season, so he probably wouldn't be willing to disrupt the four-seam/cutter/curve combination that has helped him win 10 games thus far.
A.J. has thrown 35 changeups this year (2.3%). His change is kind of like a splitter, and it averages around 88 mph. This year, only 7 have been in the zone, though 4 of those pitches have been swung at. Burnett's really a two-pitch pitcher, since his fastball and curveball make up about 98% of his pitches. Actually, I'd call him a three-pitch pitcher with one off-speed offering since he uses two distinct kinds of heaters - a rider and a sinker. But, obviously, it would be nice for him to be able to throw his change for a strike.
WRAPPING IT UP
Basically, it's the same story for all of these guys. They came into camp wanting to work on their weakest pitch, and it appears that that pitch is still pretty weak. I think that it's more important for Hughes and Burnett than for Robertson and Chamberlain, since the latter two are relievers and have more condensed repertoires to work with.
Gameday PITCHf/x data is from MLB Advanced Media; it can be easily accessed via this tool.