Shadowbanning is tyrannical. I can understand the method for actual spammers, which are mostly automated bots. In that case, the lack of API feedback is important to slowing down the bot's ability to adapt and repost. If you know it's a real person, you've got to have the deviant mind of a reddit admin to think shadowbanning is acceptable.

But it depends how you define shadowbanning. The term mostly applies to forums and comments. I don't consider curated social-media networks limiting the reach of any one node as a ban, as long as other nodes can still amplify their message. Media platforms have no obligation to recommend or trend your content. However for full transparency they should make it clear on the live/trending/search page that some content is hidden.