Nice insight. The aim of the article wasn’t to give the best approximation, we could simply calculate as many terms as we wanted and get an approximation, increasing in accuracy as the number of terms increased, if we wanted to. The aim was to show some nice summation and algebra to bound the sum between two consecutive fractions. I like your approach using integration as well!