What makes up “the story”? At the simplest level, yes, evolution will replay to the extent that organisms become more adept at gathering resources and replicating. That is pretty much the definition of evolution. But do the same niches get filled by physiologically similar organisms… well, except for very quick “rewinds”, mostly no. There is a huge body of literature on islands showing that.

How far back does this conceptual rewind go? If we go way way back, there are impact frustrations and massive geological and chemical disturbances that add a big random factor to what the starting point for extant life is. We really just don’t have the data to know if true multicellularity is a probable innovation, though it looks like it probably isn’t (a couple of billion years is a long time).

Speaking of massive chemical disturbances… there is also the little (huge really) factor of biological modification of the environment. This is the big one. The world isn’t dominated by niches created by abiotic processes, and the order and timing for the creation of many niches isn’t fixed. Imagine what would happen if efficient facilitative aerobic grazers arose a bit more quickly after photosynthesis.

Gould and too many others are fixated on multicellular life (worse really, animals). That results in a very myopic view of what evolution is as well as making it far too easy to forget just how dynamic the abiotic environment really is.

Gould’s fixation on physiological constraints makes for a very weak reasoning behind his “replaying the tape” assertion, but dynamic systems and biochemistry do provide some real indication that the course of evolution on the Earth is one of a huge (factorial probably) number of possible paths… many of which are very different by the common sense of the word. Of course, that wasn’t what Gould seemed to be thinking of.