[dbaupp]

In many ways, Excel is too flexible and yet not flexible enough for data processing: it temptingly allows people to do some analysis but can very easily build up into a crufty mess of unreadable in-cell formulas and VB macros.

Often a database, JSON or just plain CSV files, combined with R or Python+libraries, will give a far far better result (faster, maintainable, more flexible presentation, reusable, VCS-able).

[slurry]

Excel is completely terrible but it puts real analytic power in the hands of non-technical people. For most business organizations, Excel alone is worth the price of Office, even if only a couple of accountants or marketing people use its power features.

Better to solve the right problem the wrong way.

Come to think about it, "too flexible and yet not flexible enough" describes all the popular scripting languages, too - Python not least among them.