It's not really much of an argument since there hasn't been any back-and-forth. It's just different people adding their 2 cents. You'll notice the same name never appears twice in the quote tree. I'll make it an argument if you like by outlining my original objections.
My own contribution was in response to a statement that we had to have different creature types for babies because of tadpoles (presumably because of the body plan differences). I responded, no, we could already do variable body-plans with castes instead. And it would be better. Sure, this is "mere semantics" since it could be achieved either way. But there are straightforward or roundabout ways to say the same thing, and they are semantically equivalent. Some semantically equivalient ways are just better than others.
Consider the "separate creature type" implementation:
[CREATURE:TREE_FROG]
...
[BABY_CREATURE:TREE_FROG_TADPOLE]
[CREATURE:TREE_FROG_TADPOLE]
[ADULT_CREATURE:TREE_FROG]
...
etc. the baby needs to know what they will grow into, and the adult needs to know what they give birth to, so both creatures require new tags for this purpose. Now what happens if you want to copy the TREE_FROG to make a variant race? You need to copy all related creature types, and go through them to update the baby/adult creature links. Then compare that to the CASTE-based implementation:
[CREATURE:TREE_FROG]
[BABY_CASTE:TADPOLE]
[ADULT_CASTE:FROG]
[CASTE:TADPOLE]
....
[CASTE:FROG]
....
To copy this to a new creature is much more straightforward since all the baby / adult castes are relative to the creature type, you can copy this without needing to go through and re-edit all the baby types.