Steve:
I am inclined to agree with you. A typical amplifier for driving a speaker will be biased in such a way that the speaker sees no net DC signal across its output leads when there is no input signal.

The TDA7056A amplifier/driver I mentioned in the previous example is a prime example of this. When the input is quiescent the two outputs of the BTL stage place the two speaker leads at almost exactly 1/2 V++. (By the way, for any reader wanting a nice 2.5 Watt audio amplifier that can be driven from a typical 0.5->1.0 Vp-p input signal and powered by a +12V supply should look at the TDA7056A....wonderful part). In addition I believe that one will AC couple the input to this chip so that DC offset in the source signal does not mess with the bias of the TDA7056A chip.

None the less I see no real problem in driving a speaker directly in the manner shown in the circuits I presented. I have done this for years with no overall ill effects.

Michael Karas