Here is a new analysis for both arms, given the added data.

As can be seen from the above plots, the earlier model for the right arm remained quite accurate with the addition of the new data point. The predicted value for recovery for the right arm remains approximately the same as predicted in the plot I made last month. The left arm, however, was more complicated, due to the fact that the new data point did not follow the earlier model. This resulted in the linear model for the left arm being thrown off significantly, and a bad line fit (R squared value for the linear fit dropped to 0.11; very poor).

Considering the above, I tried two alternative models for the left arm, including both an exponential and logarithmic decay. Though both curves fit the data fairly well, the exponential model is quite amusing in that it predicts Virmir's left hand will never recover (obviously, this is probably not the case, ha ha!). Looking at the logarithmic model, however, we see the R squared value indicates that it fits better than the exponential one. Surprisingly also, the new logarithmic model predicts a recovery that is closely in line with what the earlier linear model from last month predicted. Given these observations, at present it appears the logarithmic model best predicts the behavior of the left arm.

More observations in the coming months!