my two pennies wrote:
My average weekly mileage has been the single highest correlating factor in my race times. A few years of mile times (including some 1500 and 1600 conversions) as a sample:
1. 35-40 mpw, 4:36 mile
2. 40-45 mpw, 4:30 mile
3. 50-60 mpw, 4:22 mile
4. 60-70 mpw, 4:19 mile
5. 50-60 mpw, 4:23 mile
6. 70-80 mpw, 4:12 mile
7. 40-80 mpw (injured), 4:14 mile
8. 80-90 mpw, 4:08 mile
9. 80-90 mpw, 4:07 mile
10. 40-50 mpw, 4:17 mile
11. 40-50 mpw, 4:18 mile
12. 80-90 mpw, 4:05 mile
13. 100-110 mpw, times TBD...hopefully can give 4:00 a scare :)
While obviously some of the performance level difference between years is down to other factors, the correlation to mileage is pretty incredible.
Taking the midpoint of those ranges and doing a linear regression says that mileage accounts for 76% of the variation in your mile times, and every additional 10mpw gives you a 4.3s improvement. It looks like you've already done that as 100-110 indeed gets you to 4:00 in this admittedly very simple model (4:02 to 3:58). Please report back...