Outlier removal in time series with large gaps

I am attempting to remove outlying values in a time series with a lot of missing values. the idea is to calculate running mean and standard deviation for 11 day window and remove values that does not fit into 95% envelope. However I only want to remove outlying values fro windows which contain minimum 3 non-missing values.

My idea was to calculate three tables - running mean, running standard deviation and something like running count which would indicate number of non-missing values for each window and then build if statement checking each value in the original data.

Have anybody attempted something similar? Does anybody have any suggestions how to implement this without the loop? Thank you