it's obviously based on the data size itself, memory available etc etc. Pretty clearly people ran the exact same test, fill a table with a bunch of empty/meani

it's obviously based on the data size itself, memory available etc etc. Pretty clearly people ran the exact same test, fill a table with a bunch of empty/meaningless data and see where performance degrades - then write a catchy headline title/post. In the end, it's more nuanced than that but i think the overall theme is know your tuning once you get to certain data sizes.