Data Analysis - Cycling Performance

I have metabolic test data (From MedGraphics - New Leaf Metabolic Cart System) that I have manually added Cadence and Power to while they were testing and on a stationary bike. A combined .xls file is attached. However, I work with a lot of clients who do not have POWER meters on their bikes, but yet they want to use those bikes to help them develop improved power. Without the ability to measure power, this presents a real challenge. As an experienced cyclists, and having done research on both Power and Heart Rate, I have a good "hunch" that there is probably a statistically significant relationship between Cadence and Heart Rate that can lead to a POWER Value. I do not expect it to be able to PREDICT power in terms of raw watts (although I would not rule that out completely), but I do wish to explore if I can estimate power in terms of Watts/subject's weight or Watts/Lb as a function of Cadence and either Heart Rate or % of T1 or T2 (specific heart rate markers) Specifically, I need to know if I can deduce what a subject's power level is RELATIVE to their MAX power, if I know their Cadence and Heart Rate, or Cadence and T1 or T2, or Cadence and % of VO2 at Thresohld, or Cadence and % of VO2 Max.