Agreed.
A small example: you have some internal database, and now different teams want to have some data analysis done.
The fun thing to do would be to write a generic reporting front end, and easy enough to use so that non-technical people can use to create their own report.
And the pragmatic solution that is to instead spew out all relevant data as CSV with a BOM, so that Excel can easily read it. Turns out a generic data analysis tool exists, and it's unlikely you can beat Excel at its job.
Producing some CSV files from a DB is orders of magnitude less work than creating your own analysis front end, so in nearly all cases, this is the solution you're paid to do.
Since I started writing software not because of any love for it, but to solve various problems I was having at the time, it's always sort of shocked me that some people manage to get so far into their careers without realizing what they're being paid for.
I quite enjoy it now, but rarely in the revel of pure creation. Usually it's from the satisfaction of a problem solved in a reasonable amount of time under well considered constraints.