`make_df.ipynb`: Repeated `pd.concat` in for loop is bad practice

[cirorandazzo]

Rather: load many df into an array and call a single concat