What's in a time?

I think I'm starting to slow down. Certainly I'm not getting any faster. I've just looked at some recent race results and compared times this year to those of 4 years ago. Take the E22/22 course: In 2005 I got round in 57 minutes 45. Last Sunday it was a measly 1:01:42. The winners time was within seconds of 48 minutes on both occassions and I could only manage a tad over evens on a (basically) flat course. Thats a loss of a minute a year. My wristwatch does better than that.
So. . . . lets check this by looking at times over the E21/10 course. The best times per year are:
2009 - 26:13 (April)
2008 - 25:14 (June)
2007 - 25:15 (July)
2006 - 25:44 (July)
2005 - 24:44 (July)
From this you could deduce that by July I'll be flying and pushing down into the low 25's again and that this years loss can be made up in 2 months. If only it was that simple and my legs and lungs could obey basic commands, but they don't, so what's really been going on? Lets have a look at miles ridden first: Up to July every year since 2000 annual mileage done has averaged 2500 with 1200/1300 done by the end of April. This year I was down to 950. Now, as moaned about in other posts, a lot of this has been due to the winter weather, so it is fair to assume that base mileage is one culprit. In 2005/6 average heartrate while racing (sticking with 10 mile timetrials for easy comparison here) was 175, max 186. This year I notice that my average has dropped to 167 but max remains at 183. Resting rate has remained constant at 52. Why is that? Is it significant? Does it explain why I'm riding slower this year? (I'm sure the mobile blogger will have the full science on that one. If not perhaps Vertical Blue would pipe up with an answer).
All of this has delayed the answer to the question posed in my last post - Is a disc faster than spokes? But it has lead to a conclusion on tri bars! Lets look at some club 10 data again.
Here are the best times on fixed:
2009 - 26:40 (May)
2008 (None ridden)
2007 - 26:14 (July)
2006 - 27:17 (June)
2005 - 27:12 (May)
We will notice that this year the difference in times is purely seconds, but in previous years around 2 mins slower was to be expected. (Are you still paying attention at the back of the class?) For 2009 I made the following changes to the track iron:
1 - Tri bars.
2 - Disc wheel
3 - Switch to a plastic bike with a lower front end.

As I've used the disc consistently all year this year, I think we can eliminate that as the big improver. Likewise fitness. The Dolan plastic (sorry - Custom laid carbon fibre monocoque frame) is actually heavier than my trusty 531 Bob Jackson, and certainly not as comfortable to ride as shown by the fact I've stuck to the Jackson for the track league. That leaves tri bars and leads us to the reasoned assumption that it is they who account for such a spectacular closing of the time gap. Does anyone know how this conclusion compares to manufacturers claims or any proper test data?