if he's at 70rpm the motor isn't giving maximum assistance.... yes.
So if he changes into an easier gear, he can spin his legs and won't be personally putting in any more power, if he spins a bit faster with the same power, the motor will give him more assistance, and he'll go faster, with no extra input from himself.
The graph you posted earlier is misleading you and some others because it shows output power, not torque.
Your opening statement that the motor isn't giving maximum assistance is sort of wrong, but it depends what you mean by assistance. If you mean power, you're correct, but to get the motor to give it's maximum power, you have to give a greater proportion of the total power yourself than you would at a lower speed, so proportionally, it's giving less assistance.
From the graph:
At 70 rpm the power is 540w
At 85 rpm the power is 570w (maximum)
Rotation x torque = power
we therefore have:
70 x T1 = 540, so
T1 = 7.71, where T1 is the torque the motor produces at 70 rpm
85 x T2 = 570, so
T2 = 6.71, where T2 is the torque the motor produces at 85 rpm
You can see that the torque is significantly less at a cadence of 85, so the only way to get the motor from 70 rpm to 85 rpm is to put in significantly more torque yourself. That assumes that you're already in bottom gear. You'd have to cover the 13% missing torque from the motor plus 15/70, which is another 21% ,to get to 85 rpm. That means that when hill-climbing in bottom gear,
to go from 70 rpm to 85 rpm, you'd have to pedal 34% harder.
If you look at a graph of the power a person gives when walking, trotting, running and sprinting, i.e. speed vs power, it'll show that they produce more power at high speed, but that doesn't mean it's easier to sprint everywhere.