Clean up docs language

[thomasteisberg]

Removes some claudish that snucks into the docs, fixes an equation that GitHub markdown wasn't rendering correctly, and clarifies the QC filter

[elliesch]

This looks great! Thanks for sharing these changes! Would it be possible in this paragraph:

As of 2026-09-03, the input data is filtered using QC flags that identify surface saturation and image combine errors that could lead to radiometric miscalibration. The filter drops about 14% of Antarctic and 13% of Greenland traces but leaves the model fit essentially unchanged. Scored on the same held-out points, the pre- and post-QC posteriors differ by 0.03 dB in RMSE and the predicted maps move by less than 1 dB almost everywhere.

to add that when the predicted maps move by <1dB, what ratio of the overall signal this motion is? I.e. for each point that has moved what is the residual/signal for that point, and then what is the average residual/signal across the new map?

[thomasteisberg]

Sure! What do you think would be helpful? The model's built-in uncertainty (sigma) has a mean of 12.7 dB. What do you think would be a helpful way of putting 1 dB in context of this?

[elliesch]

Thanks @thomasteisberg! Would it be possible to see a map with measure of the difference (~1dB or less) over the measure of the estimated observation (the mean of the estimate rather than the SD)? I'm interested to see if there are any regions where the difference is large, or if overall it's small compared to the signal