Test E2E Chain hashing on Butterflynet

[masih]

After #792 is done, run an end-to-end test of F3 on butterflynet to understand:

• any gaps in integration between the new chain hashing and Lotus in terms of e2e flow.
• impact on reducing bandwidth consumption
• correctness of design/implementation overall.
• sanity of default chain exchange parameters, such chain rebroadcast interval, pubsub settings, default Manifest parameters.

This should help de-risk unknown-unknowns of the major changes introduced by #792.

Note that there is nothing blocking us from doing this as soon as the code is ready.

We may need to increase the size of butterflynet for better testing. But as initial step a minimal butterflynet setup should help cross off a lot of the boxes in terms of correctness of implementaion.