The monikers "boost tab" and "servo tab" are sometimes used to describe the same thing - a bit confusing. I wrote an article in MAN about both boost tabs and servo tabs about ten years ago. Boost tabs move passively in the opposite direction to that of the control surface. For example, when an elevator angles up, the boost tab angles down in relation to the elevator. This causes it to produce an upward aerodynamic force near the trailing edge of the elevator that assists the servo, or muscle power, in moving the elevator up. The boost tab normally does not move to a larger angle than the control surface deflection unless the tab is a bit undersize. I always adjust the linkage of the boost tab to produce less angular deflection than I expect to need, then gradually increase it until I get the control response that I want. Too much movement of a boost tab can result in the control surface overpowering the servo, and forcing the servo to restrain the movement of the control surface. Evidence of this is that the model just refuses to trim, since the control surface tends to flop up and down within the linkage slop. I find that boost tabs can reduce the force needed from a servo by up to about 85% before the system gets a bit 'twitchy', or non-linear.

The term "servo tab" normally applies to a tab that is controlled directly by the servo, with the control surface free-hinged, and mass balanced, to weathervane in the slipstream. Some large airliners use this system, which provides beautiful control feel and permits muscle force alone to control a very heavy airplane. Examples are the DC-9 and MD-80 ailerons and elevators, and DC-8 ailerons, which use servo tabs connected directly by cables to the control column, and boost tabs on the control surfaces. The combination probably provides a control boost multiplier of something in excess of 100 times.

I am not familiar with the term "offset tab".