I seem to recall an article from Joel on Software about the same phenomenon from a procedural point of view, where he compares increasingly complex institutional reaction times to the increasing amount of ritual that is being performed before we leave the house (with old people taking the longest time, because based on past experience of all things that went wrong in the past, their checklists before leaving the house become absurdly long over time).
Just an interesting side observation about the VLC speed thing: my personal sense of rhythm has always been so bad that I couldn't tell the difference between 100% speed then and 80% speed now. But an observation I recently made was that I can now comfortably listen to videos at 110% or even 120% their speed without losing content - something I certainly couldn't do at a younger age (in fact, I remember I often had to pause recordings in order to think about stuff that just happened or even just to empty my "language recognition buffer" before I could move on). So while it may well be true that 1 second doesn't last as long for me as it used to, I certainly seem to have become more efficient at processing some types of data.
If you don't want to do those experiments yourself maybe you should eMail some Professors who are interested in this topic.
Anecdotally, when I was 8 or so I once sat there for an hour or so memorizing the rate of the clock (so I could later mentally estimate intervals of time without a stopwatch). At the age of 23, I noticed and then calculated that a second in my "internal" clock was about 1.3 seconds. Since then I have used VLC to re-watch movies I saw as a child at ~80% speed. The musical rhythms seemed closer to what I remember.
Obviously this is insufficient support by itself, but I's love to see this phenomenon studied more rigorously.