One good reason might be more data ownership than tracking. Centralized platforms seem to at some point realize how valuable their user’s contributions to the site are, and change the deal eventually to e.g. lock things behind a paywall when they grow bold and/or can get away with it. Reddit and Twitter are two high-profile examples of this, but there are numerous examples from the past as well for things like nutrition tracking and cd cataloging. A federated model might better reflect the idea of community ownership, and prevent site owners from betraying the community and cashing out on their contributions.
You can still export all your Goodreads activity to a .csv file. Amazon has so given up making any improvements to the website, that I don’t even expect them to remove functionality like that.
I don’t believe that Amazon considers users’ contributions to Goodreads valuable, it only bought the site to 1) stop paying Amazon Associates referral fees to the independent corporation that Goodreads was, and 2) to stop Goodreads from directing any book-sale traffic to its competitors.
tbh, this is not my issue. My only complain with Goodreads is their pathetic UI. My own purpose of using Goodreads is tracking and the UI is so confusing that it takes me a while to figure out how to find the last page I had read and then update it.