This point isn't necessarily true, depending on the goal. If the goal is to finish as highly ranked (in NW) at the end of the reset, it is a smarter play to take land from people you are already ahead of, because doing so costs less.

How does this cost less? Simple. You might spend "more turns to take less land", but at less risk of losing said land. Using more turns in this case isn't actually a downside, all my excess turns go into exploring or cashing anyway, since income generates every turn (except for techers). Being retalled means you lose CS, you lose tech, you lose a portion of built land, and attacking someone larger than almost always requires more jets and oil than attacking multiple small countries, such that this loss negates most of the potential gains from the landtrade, unless you were attacking someone 2x-3x larger in landsize. Even if it was profitable to do so compared to "bottomfeeding", you would eventually reach an equilibrium point where it starts to be better to bottomfeed than to topfeed, as you approach the land sizes of the larger countries.

The larger your country is, the worse topfeeding becomes. At the extreme example, the biggest country (in land) cannot topfeed, since any retal would not be profitable at all, and such a country must bottomfeed.

This is why it is often explained that "if you have to topfeed, you are already in a 'behind position'" to begin with, and a good player doesn't let himself fall to a "behind position" in the first place outside of external factors such as being suicided.