I was somewhat skeptical myself. However, you can reset to factory and then listen to a few things... then use Trueplay to do its thing and listen to the same things. I definitely notice a fairly meaningful difference. I didn't expect it to be dramatic but I would claim dramatic to my ears and for my purposes (which is mostly music).
Of course, in a "blind" test, someone could manually change EQ and fool someone that a Trueplay-type process was run. But the idea here is that maybe a program can "listen" within each owners room to tweak settings to try to make these speakers sound as good as they can in that room.
This kind of idea has been around for a long time. A receiver I bought about 10+ years ago came with a long wire to a mic that was supposed to be put in the prime listening position to auto-adjust my speakers to optimal listening from that spot. iPhones now claim to do AppleTV video calibration for TVs (usually a multi-hundred dollar service by an in-home tech). HomePods claim to do the same for audio. I give Sonos some faith (too) that they can actually squeeze more out of their speakers by this process. Else, my discerning ears are effectively fooled by using it.
Given that it is FREE, even if it is a placebo, Sonos owners can easily give it a try themselves and hear or not hear a difference. No big loss either way... but possibly a gain- imagined or actual- like using an iPhone to adapt AppleTV video quality... and all of the "magic" we mostly accept accomplished by HomePods.
There a number of similar, and arguably more sophisticated system available for other speakers too. I would hope they are more sophisticated as they cost more than a Sonos play 5 or even in some cases a whole house system. Sonorworks, Genelec GLM and Trinnov to name a few. But they all work in a similar way to some degree - there’s a microphone to sample the frequency response of the room, a box of tricks that calculates driver time alignment, crossovers, some eq and way to counter any peaks or dips in the desired frequency response curve - which is usually aimed to be flat as most of these systems are used in high end mixing and mastering studios.
So whether subtle or a mild placebo. Considering it is free, built in and a breeze to use then can’t complain much about True-play*.
I’ve over 25 sonos speakers inc two 5.1 systems in the house, the difference true play makes is mostly down to the room and positioning. But mostly it improves the overall performance, optimised for placement. When you have a two sub system it also manages that as well as adjusting crossovers with the Arc.
*There will always be a few entitled c***s who have little experience of it and expect it to do the work of a $10k Trinnov system on their tiny $200 speaker, but they are a special breed best kept in their moma’s basement.