Do you know a proof for that? I.e., that for n logical qubits, decoherence time goes down polynomially rather than e^(-a n)? AFAIK (and yes, I could be mistake

Do you know a proof for that? I.e., that for n logical qubits, decoherence time goes down polynomially rather than e^(-a n)? AFAIK (and yes, I could be mistaken!) what we deal with is reducing a susbtantially.