Both conjectures start from the same diagnosis — phenomenal binding of distributed neuronal feature-processors is classically inexplicable, and neither discrete decohered neurons nor mere synchronous firing can generate unitary experiential objects or a unitary phenomenal self (Pearce, "Schrödinger's Neurons"). Both also assume, per your premise, that the superposition principle never breaks down: no collapse, no objective reduction. But they locate the binding resource differently.

(1) Superposition of neuronal feature-processors ("Schrödinger's neurons")

• Neurons are not effectively classical, discrete, decohered units. Distributed feature-processors enter quantum-coherent superpositions — individual neuronal "Schrödinger's cat states" — at sub-femtosecond timescales (physicalism.com).
• A moment of phenomenally bound experience just is one such coherent macroscopic neuronal superposition: a fleeting cat state is a phenomenal simple, not an aggregate or classical ensemble of anything more primitive ( hedweb.com/physicalism).
• The components of the superposition are the feature-processors themselves (edge, colour, motion processors, etc.); the single global state-vector with its non-zero phase relations and interference structure is the bound percept. Selection pressure harnesses sequences of these coherent states to run a fitness-tracking world-simulation at roughly 10¹⁵ coherent "frames" per second (physicalism.com).
• Decoherence does positive explanatory work: thermally induced dephasing is progressive phenomenal unbinding — it explains why the mind "becomes unstuck" so fast, why binding is evanescent, and why neurons appear quasi-classical to neuroscientists in the first place ( hedweb.com/physicalism).

(2) Entanglement-focused conjectures

• The proposed binding resource is nonseparability between subsystems: entangled states (e.g. across microtubules, ions, or spins in different neurons/regions) are invoked as the physical "glue" that links otherwise distinct micro-experiential units into one mind ( The Unfinishable Map).
• The emphasis is on entanglement as a relation: two or more systems share a joint state whose statistics cannot be factorized, and this nonlocal correlation is supposed to constitute or underwrite phenomenal unity.

How they differ

The core philosophical difference: entanglement is a measure of correlation between distinct systems, whereas superposition makes possible one object with many internal features. If what needs explaining is how the redness, roundness, and visual location of a tomato are features of a single experience — not three correlated micro-experiences — then correlation-type resources arguably can't deliver union. Two entangled particles remain two; a superposition of feature-processors is one state. Notably, an entangled state is itself just a superposition of product states, so on reflection the unifying work in any quantum-binding story must be done by the coherence of a single global state anyway — which is exactly what conjecture (1) takes the bound mind to be, while entanglement conjectures risk treating a statistical/structural property of states (nonseparability) as if it were phenomenal glue (Pearce, "Schrödinger's Neurons",  hedweb.com/physicalism).

Both faces of the family share vulnerabilities — chiefly Tegmark-style decoherence timescale objections in a warm, wet brain — but conjecture (1) additionally predicts something entanglement conjectures do not: that interferometry on living neuronal tissue should reveal coherent, functionally harnessed cat states, making it directly falsifiable rather than merely suggestive of quantum correlations (physicalism.com).