RMS should not change in meaning between different applications. But in some situations, people may "cheat" a bit when using the term.

As noted earlier, the piezo speakers are capacitive, which means that they will not swallow all energi that an amplifier may be ready to send out since the piezo speaker will not have any 4 ohm or 8 ohm constant resistance.

That means that you may have an amplifier that emits 75 W into a 4 ohm resistor for the full supported frequency range (say 40Hz to 80kHz). When you then replace that resistor with a piezo speaker, the output power will increase with increasing frequency. At 40Hz, the impedance will be extremely high, so the resulting power will be almost zero. At 80kHz, the impedance will be quite low, so that piezo tweeter will most probably die.

That was what I meant when I linked to a number of tweeters and noted that they did mention 150W dynamic power and (some of them) 75W rms. They would probably not survive 75W rms if you did try to feed them 75W at 80kHz. But the manufacturer selected to "pretend" that it would survive the same voltage from the amplifier at 10kHz as if you had used a normal coil tweeter rated at 75W rms. So they may kind of "cheat" a bit with the rms specification. For a practical purpose, a small 30 gram tweeter can be seen as similarly capable as a very large and heavy, oil-cooled, coil tweeter just because the piezo tweeter is self-limiting the amount of power it will "pick up" within the audio range.

Because the piezo tweeter have extremely high impedance at low frequencies, you can normally connect them directly to the amplifier without any filter - you don't need to remove the bass or middle register frequencies because the tweeter has so high impedance for the energy in these frequencies. But since many modern amplifiers have a linear power frequency range up to 100kHz (with power frequency range, I don't mean the frequency range at 1W @ 8 ohm but the frequency range at the rated output power), the amplifier are capable of emitting enough energy to quickly kill the tweeter in the frequency range where the tweeter impedance is low. To protect from that, you can use a series resistor. An 8 ohm series resistor will not only reduce the maximum power the amplifier can emit at higher frequencies but will also reduce the risk of any high-frequency feedback loops. After all, no music material has any sound at 100kHz - a normal CD can't encode anything higher than 22kHz. A DVD can reach higher but you still have filters after the DA converter. The higher sampling rate for a DVD just allows a less sharp filter curve to be used, and hence less problems with phase shifts.

No, I haven't really seen any high-power audio piezo speakers. For audio, you normally use many tweeters to improve the total power rating. It isn't uncommon to see large PA speakers with four piezo horns - this also allows the direction of the sound to be controlled. Now and then, I have seen interesting special transducers intended for quite high power. But while they are fun to read about, I don't have any practical interest in them, so I haven't saved any links for future reference.