Decide and integrate the blob module with authz

[evan-forbes]

We are adding authz to the application in #1540, but we are pushing the integration with the blob module to a different PR. This is mainly because there are different ways to do this, each of which unblocking potential features.

We should decide on and then implement that integration. The decision might involve an ADR.

[liamsi]

Also, everything beyond the generic authorization should be accompanied with a test.
re:

• feat(app)!: add authz to the state-machine #1540 (review)
• Integration-test re governance & feegrant module #1541

[cmwaters]

@evan-forbes Is this something we want to punt till after mainnet?

[evan-forbes]

its less important than other things imo since we have feegrant, but if we can we should try to squeeze it in per the onsite decision.

[evan-forbes]

to update this after a sync discussions. Since authz requires wrapping the PFB, this complicates the block encoding, specifically for zk rollups or custom parsers of the PFB namespace.

we are also able to get the same functionality using a custom feegrant interface per investigation by @cmwaters, so I'm going to remove this off of mainnet but won't close the issue.