It looks kinda messy. For a human DM this seems quite horrible to track.
For a human DM, they'd be using RAI. In which this interaction doesn't occur.
RAI states that only the damage gets transferred over and no additional effects from the original damage source, with the only interaction being the "Double Dip" of mitigations because of the two pronged instances of damage still working on regular damage rules (I.e. Damage received is the net result of incoming damage modified by resistances/vulnerability then modified by negations)
The quirk in regards to "It becoming a spell from the target of the bond" is just spaghetti code. Wherein because the source of the new "Damage" is technically the bonded target (Because the instance of damage that creates it occurs from the target receiving damage) then it starts to get modified by that allies effects as if they cast the damage spell. Essentially, Larian forgot to code in rule exceptions for this damage to not benefit from character modifiers (Which it shouldn't anyway, not even if it was sourced directly from an enemy as Warding Bond only transfers damage, not additional effects)
This quirk shouldn't exist. Especially since even IF the separate instance of damage was supposed to be "As if cast from the ally" then it'd be based off the Cleric's modifiers since THEY cast the Warding Bond spell that is dealing the damage, not the bonded target.
But yeah, in 5e and how it should work in BG3, is simply take the damage bonded target received (Which is what it took after mitigation) and deal that to the Cleric (Who then receives damage after their own mitigations).
