Does clean power matter?
Yes.
Does it matter enough to take steps to get cleaner power than you already have?
That’s a much more nuanced question with no simple, reliable, answer.
For a start it depends heavily on the gear being used. And more so on both how “dirty” your power is and the nature of the “contamination”. And these factors, even if significant enough to warrant doing “something” about them, generally require different types of remediation to fix effectively.
Switch-mode power-supplies, good ones operating at very high switching speeds at least, are generally all-but immune to voltage fluctuations and line-noise at most frequencies. Putting filtering ahead of those rarely makes a usefully measurable, much less audibly-discernible difference (absent a literal fault in the supply).
However, that same SMPS might spew enough ultrasonic noise back onto your AC line to affect the linear supply in an amplifier. So what? You might think. It’s ultrasonic … you can’t hear it. True … to a point. But if it’s measurable, even at levels that are below the threshold of audibility, then your amplifier is still going to try and amplify it, and not all amps are stable (can oscillate) with ultrasonic noise, so that can cause very audible issues in the audible band and at levels that are high enough to hear.
That’s a simplistic example, but a real one.
How well designed and filtered the PSU on your gear is will also make a difference. It might be MORE effective for its intended purpose than anything you can add before it - potentially rendering upstream treatment moot.
Having things like your fridge on the same circuit as your tube amps is probably a bad idea (for multiple reasons). Those nasty PowerLine network adapters, that use your AC wiring to transmit data, tend to cause issues too. Not necessarily audible, as that again depends on many factors, but certainly not desirable.
Absent a way to measure and evaluate your AC quality directly, and the effects of different “treatments” at the outputs of your gear, then you’re down to listening tests and switching different types (or combinations) of filter in and out. And in that vein, if you don’t hear a clear difference with filtering/conditioning in place, I’d spend the money elsewhere.
