Sort of. I think it strengthens "the user is free to run the code", to implying that a user who has that freedom (to run the code) - has the same rights as someone who got a copy of the binary. To (be able to) ask a server to run code through an api, is seen as similar to run code on your own computer (a device, which might not in fact control either, hardware being what it is). It highlights the fact that the GPL is for users. If you use a webmail service, you should be able to fix issues with the code - just as if you ran a desktop mail user agent.
So sure, it changes the definition of "distribute" as it is used in the GPL - but my understanding of the "why" is to to be explicit about what is really needed to guarantee users the four freedoms (run, study&fix, copy, share fixes).
So sure, it changes the definition of "distribute" as it is used in the GPL - but my understanding of the "why" is to to be explicit about what is really needed to guarantee users the four freedoms (run, study&fix, copy, share fixes).