docs: Add 2 chains to feature matrix: fuse & blast

[p-diogo]

Adds full protocol support for moonbeam, fuse, and blast data sources.

The Council will be reviewing the CIP reports for these chains this Thursday.

Next steps:

• Create a GGP referencing this PR (Council approving full protocol support for these 3 chains)
• Wait for the Council to vote
• Add GGP to the feature matrix
• Merge PR

EDIT: removed moonbeam from the list.

[abourget]

As discussed recently with @mangas .. there's an issue with enabling Substreams-powered Subgraphs for names like "moonbeam" is tricky because those chains do not have Extended traces instrumentation..

This creates a problem of data provenance (the source backing a subgraph) needs to be holistically defined, and if we go from Light blocks to Extended blocks, all under the "moonbeam" network name, things will break determinism.

The graph-node is able to abstract from one to the other (offers triggers from both sources), but Substreams get fed the big raw blocks, and those need to be deterministic.

I think it's not safe to enable SpS without defining the exact meaning of moonbeam, and then be ready for, in the future, support moonbeam+extended (and consider moonbeam to always be light blocks).

[p-diogo]

As we're still debating how to best handle different block types (base or extended, related to @abourget message above), I'm removing moonbeam so we can proceed with the other non-contentious chains.