Pandas:using list as cell value – different approach to modify values and NaN issues

This is 50% a question and 50% an observation that baffles me a bit. Maybe someone can enlighten me.