The more data you bloat API endpoints with the more overhead there will be, and if more internal services have to be used that also adds more delay to the requests. Additionally because the data may come from multiple services internally that could also mean more dev time to develop and maintain those systems. Your train of thought so far seems to think that it magically takes no effort, time, or overhead, to add more data to the API, so I thought I’d try to clear that up a bit.
Because as you yourself agree that different people need different things that just shows that some data will be unused on requests. This is also backed up by how v5 operated where there were fields that were so niche they usage couldn’t justify the overhead added to ALL requests to those endpoints.
And again, this is exactly why UserVoice can be so valuable, so Twitch can see the demand and use cases for adding new features and data and weigh that usage up against the increased overhead, dev time, and maintenance.