The universe doesn't care what sections of it we consider to be organic or inorganic, one creature or multiple creatures, alive or dead, one species or another, sentient or not. In the end, they're all leaky abstractions.
Imperfect as they are, these abstractions impact the universe though, through our thoughts and actions, and getting them wrong can have dramatic consequences.
Thinking in terms of composable attributes maps the world a lot better than dichotomic, essential taxonomy (composition > inheritance in the CS world).
Taxonomy is a premature optimization that comes intuitively because it's been selected by evolution since it was good enough for most purposes.
I'd not be surprised if we learn in $time_unit that everything in the universe is life. The variable is - like in relativity - the experience of time, not if something is alive (as opposed to moving in relativity). It may be close to zero from our perspective, so quite hard to spot, but we get better measurements with more time.
I don't recall Spolsky's article (even though the term rings a bell), but as someone having worked in modelling and simulation the better part of my career, I tend to think of it more along the lines of Asimov's "Relativity of Wrong": every "thing" is a model, and no model is perfect, but some models are better than others. If you're waiting for the perfect model, you'll never do anything, which brings to mind another adage, adapted: strong models, weakly held. That is to say, do the best with what you've got and don't be afraid to change once you find a better model, but always remember that they are models with limits.
The fine article linked here about fungi I applaud as improving our models.
https://www.joelonsoftware.com/2002/11/11/the-law-of-leaky-a...
The universe doesn't care what sections of it we consider to be organic or inorganic, one creature or multiple creatures, alive or dead, one species or another, sentient or not. In the end, they're all leaky abstractions.