Handling .txt files can be problematic with strings that are treated as numbers having zeroes deleted and changing the value. If we use .csv the value may be trimmed or decimal places removed but the value is retained and can be expanded - therefore more useful.
@AmandaBirmingham may not want this because the notebooks expect .txt - can we confirm with her that this won't be cumbersome to encode for all notebooks in metapool?