That's how it works right now with blocking characters as well. If you block somebody they disappear in your client only. The server is still sending all their data, chats etc. to your client. Your client is only blocking it by account id which is the whole reason for the privacy scandal.
If SE add this, they will do it the exact same way as they did it before. Meaning: All blocked clients will receive a message: Hey client, this user blocked you so dont show it to the person in front of the screen. Rendering it absolutely useless and easly bypassable.
This also goes for you. SE has proven that they will implement a feature like this exactly like I have it outlined and the reasons for that are to be found in their server-side backend code. You are lecturing a programmer with 20 years of experience here. They simply don't have the ability to filter traffic going out to clients as they are using a cell based synchronization in a pubsub way. That's exactly why they broadcast the account ids now to make the "block one character and block all alts at the same time" function work.
Changing this would A) take the will to refactor their core synchronization code and B) cost more in upkeep due to increased resource requirements and they have proven that they are not willing to do so.
And this idea in itself is flawed and destined to fail as long as there are plugins.
Even
IF the server stops sending the data to the client a simple check with another client whose account isn't blocked via a third party tool will reveal the fact that they are blocked. Same goes the other way around.
So the whole concept is insecure and privacy invading and the way SE implemented it is the worst way possible.
EDIT: I do recommend taking a look into this thread
https://forum.square-enix.com/ffxiv/...n-circumvented