This suggestion creates tables that are easier for humans to read, but harder for machines.
Novices are likely to perform layout in a spreadsheet program (e.g., MS Excel), but the resulting files are difficult to work with. Many government data sets have this problem: agencies will offer a downloadable Excel file that is human-readable but more difficult to load into R or Python.
Tables that look like this on the web are harder to scrape.
I ask that authors provide both human- and machine-readable formats for all data sets.