That surprises me.
For two reasons
1) The whole UK ethos of the site as an educational tool depended on people being able to find responses to their observations, identifications, and projects. Casual users may have come back a few days later, and looked at their observations or few, but for anyone using the site more intensively a more efficient way is necessary. I don't want to have to look at a dozen observations once a day for say a week, and interactions occurring after a relatively short term would never be seen.
2) They should have seen the usage of CHANGES in the server logs. Perhaps they failed to distinguish between the usage patterns of guests, casual users and power users.
We don't necessarily bring need CHANGES as it was implemented. If they brought back UNREAD with suitable filtering (by user, to include identifications and comments, or alternatively by observer, by identifier and by commenter, and by group and by taxonomy) that would serve the same purpose. (I used to use CHANGES with everything by observations where I agreed with an identification switched on, and switched on the last once a week of so to pick up later identifications and comments. I used FAVOURITES to add vascular plants that I couldn't identify so that CHANGES picked them up.)
Before UNREAD lost filtering by group I used to read every plant observation, identification and comment - I might get behind, but I'd catch up on a wet day. With filtering by taxonomy (and some substitute for changes so that identifications and comments are seen) fern or moss specialists could log in once in a while, and go through the observations that had built up. (Or I could ignore moss and liverwort observations for most of the time, and every now and again use a different filter to look at them - I'd suspect that it's more educational to process them in batches.)
Summary: I think that a sufficiently powerful UNREAD is more flexible than CHANGES, and you can put in a CHANGES link which resolves to UNREAD filtered by user.