I agree with @Dimitar. Specifically, “crypto miners” in websites is almost a meme in how bad of a reputation it has. Compensation is only part of the issue, overloading an already bloated platform (web browser) is another.
I would personally rather see Safe proper (nodes) gain the ability for low-latency operation. Livestreaming is just beginning to work its way into p2p network architectures, Safe should be able to take design inspiration and do an even better job. Perhaps there could be a loose parallel with bitcoin and lightning where stateful data sync takes some time but ephemeral data sync happens faster. Data resilience on one side, low-latency on the other.
When you use torrents you can set your own seed ratio right, lets assume the Safe Browser by default has seeding disabled. However, when you increase your seeding ratio more peers will want to connect to you and exchange chunks that they don’t have and you don’t have for mutual benefit. It’s completely up to the user what they need/want and for sure we should never mandate additional data usage of the user’s internet bandwidth if he doesn’t choose to opt-in for this.
Regarding to copyright issues and peer-to-peer seeding, if all data is encrypted there is no way of knowing what someone is sharing. It’s literally unreadable data, only accessible to those who know they are receiving and assembling it as far as I understand.
So unless they want to completely ban peer-to-peer encryption which would be really hard considering it’s very difficult to even identify an encrypted stream of data (what it means, where it goes, etc). Remember the streamer in this case is already paying to store his stream onto the network to all the farmers who will eventually store all chunks. Farmers don’t lose any money as the data will propagate and be stored nonetheless. The network however for reading and transmitting the data could have a massive boost in the increase of speed by all those peers joining to access that specific hot data on the go.
This is a bad idea and intimate browser integration is a messy interface. KISS.
Cache nodes with loose coupling are a better and cleaner solution. KISS.
For example, if a cache node was operating on the local machine, then just have the browser’s cache folder point to the cache node’s storage. Any chunks that were GET by the client would be copied to the local cache node. The cache node could then later serve the local client or any other network request for those chunks.