Maximum Theoretical Speaker Efficiency

Going with Gs5556 theory of 10% efficient and say, 93 dB. Each doubling of power equals 3dB increase in efficiency. so that would be:
20% 96
30% 99
40% 102
50% 105
60% 108
70% 111
80% 114
90% 117
100% 120dB

My logic is probably pretty messed up BTW.

Speakers are typically 5 to 10 percent efficient. You can never get 100 percent of the electrical energy into sound because the driver cone to air impedance coupling is so mismatched that it cannot be a perfect "transformer". But the other factors involved - the mass of the driver, magnetic energy, ports, box construction, etc. - all vary from speaker to speaker. For example, if a speaker is 90 dB at one watt at one meter, and it is 10% efficent, if all the power went to the speaker it would be as if the power was multiplied by 10. A power multiplier of 10 results in 10 dB or 100 dB/W/m. But other speakers have differing sensitivities due to different drivers, crossovers and construction, so there are tons of theoretical answers.

It's similar to asking: how fast can a car (about 20% efficient) go if all the energy in gasoline is converted to velocity. You would have to establish a base weight and ignore heat, braking, steering, electrical devices, lights, friction and other physical quantities. Cars have different weights, so the answer depends on a given weight, of which there can be an infinite possibility.

Assuming all the power went into sound power with no loss from the speaker itself wouldn't maximum sound pressure be the maximum levels that can be produced given our atmospheric pressure...can't remember the exact level but its something like 180db.