It's likely most of the acoustic effect is due to the impulse response of the filament going through a rapid temperature change. A filament has a positive temperature coefficient of resistance, that is, a lower resistance when cold and higher resistance when hot (that's why they burn out when you first turn them on), and also a positive mechanical coefficient in its dimensions. As the filament achieves steady state temperature after you turn it on, it expands in size generating a mechanical force. Hence, a sound.
I still like the magnetorestrictive theory as well. There's probably a little of that too.
I still like the magnetorestrictive theory as well. There's probably a little of that too.