I half agree that technical feasibility isn't everything. For example, I can murder someone with a knife. But it's not what we want as a society.
But it's not never. For example, I can see your post. So I can send a screenshot of your post to my friends to dicuss it. I can't see your hard drive contents. So I shouldn't hack into it and send a screenshot of your hard drive to my friends to discuss it.
So technical feasibility influences what is reasonable to do as a society. It's not "feasible = reasonable", or else murder would be reasonable, but it definitely does influence it.
And in this case, I believe what the Fediverse people commenting in the GitHub issue want to be unreasonable. It is unreasonable to publish something publicly, on a federated network, where privacy controls exist (but are not being used), and then claim to have an expectation of privacy in a public space, especially when such bridges provide utility and benefit to others that just can't happen if it's opt-in.
It is reasonable to block it. It is unreasonable to expect everyone else to restrain themselves from using public data in the spirit of the open Internet. It is especially unreasonable to harass non-profit bridge developers in this case. That's not a social solution, just harassment.
Note about copyright: that's not a path one should go down, because it'll make the Fediverse illegal as a whole. It's probably fair use, anyway.
Note about AT protocol: yes it's designed by a for-profit, but it's good. Just because something is for-profit and VC-backed does not mean it will enshittify; take Element for example. It solves a lot of issues that people were having with Mastodon such as global full text search and a global feed. I would use it if it only had more relays to spread out the control.