I use a lot of effort to reveal the role of the common emitter current source (the "long tail") in the differential pair to my students.
I do this by first justifying the need for a (whatever) resistor... then considerably increase the value of its resistance... and finally replace it with a current source...
But to find out if they really have managed to figure out its function, I provoke them by this question... claiming that there are (at least two) situations where the simple resistor plays perfectly the role of a current source.
Are you agree with me?