It looks like Golden Cheetah uses a quadratic regression curve between CP2 and CP180 to estimate mean maximal power. So, when you significantly increase a max at one end, it is natural for the other end to fall. I would guess that your theoretical CP2 to CP6 values are higher than before. This makes sense from a general perspective: someone who is significantly improving their fitness at high power levels may see some fitness tailing off at lower power levels. Finally, it’s possible that the regression is weighted more heavily at certain ordinate points (i.e. zones like CP2,6,30,60, etc.) and you may have hit very close to one.

As far as commentary, it looks like the updated Cheetah curve is a much better fit for your data. I know you race with your power meter and have LOT of data, so I would guess most of your mean maximal powers are very close to realistic for you (even up to CP120 — see breakaways). I think it’s better to work off past data — when that data is efficacious — than off theoretical max powers anyhow.

Frank

p.s. I enjoy reading your blog. Keep it up!