It seems heavy than maintaining something like Reader, but perhaps that's incorrect.

Very very little at Google survives on autopilot. There’s a constant drumbeat of required changes, whether they be accessibility, regulatory (DMA, hurrah), internally motivated changes (material design++) etc. For anything to survive there must be someone who cares and is willing to invest in it. This is a large part of why stuff gets killed, the default there is for things to die. Yes there are strategy changes too, but very often things die because the person who could champion a thing is no longer there to do so, or has changed roles significantly enough that they can’t plausibly do so anymore.