Hi Ken-ichi
I still don't see why users should have the choice to opt out of taxon changes. If a largely unified taxonomy is something that iNat tries to achieve (by following external taxonomies where these exist), and thereby facilitating data exchange with GBIF, EOL and others, then I think allowing users to opt out of taxon changes is leading to nothing but ambiguity. If I take your argument further, then every iNat user should be allowed to use whichever taxonomic treatment he/she prefers, and that doesn't seem to be current iNat policy for good reason.
Apart from that, I'm concerned what will happen to IDs affected by taxon changes over the years and decades to come - there will be an increasing number of inactive users, either because they've left iNat, or because they are decomposing. If the system is left as it is, observations owned by users who have opted out and affected by taxonomic splits will be accumulating nonsense.
I'm all for informing users of taxonomic changes that have been made, and how their observations are affected by that! I just don't see a good reason for the 'opt out' choice.
Jakob