In any Excel sheet everything is mixed up: you got raw data, calculated values, formulas, formatting, plots etc. With a Jupyter notebook you have these well se

In any Excel sheet everything is mixed up: you got raw data, calculated values, formulas, formatting, plots etc. With a Jupyter notebook you have these well separated, and there is a little chance that someone somehow changed the font in one of the 20k cells with values. Scaling up plots or changing the color scheme are explicit: you must have a cell changing the defaults. Yu can not just muse over some cell, press Del and plonk it. If you verified say md5 checksum for the raw data file it contains what you expect it to contain. If you follow up actions of an average Excel user, then it is a long chain of mostly purely manual manipulations, like selecting the cells, copy and paste, add color here and there etc. A lot of steps to check and verify. Compared to SQL or data frame manipulations it is too complicated to be really useful IMHO