Adopt go.work, rename modules to prep for monorepo

[protolambda]

This PR:

• renames the modules to their final destination name in the monorepo
• splits the packages into modules: op-node, op-proposer, op-batcher, op-e2e
• moves ops to ops-bedrock, to prepare for monorepo
• Many Makefile and Dockerfile updates to make the above work
• updates CI (including getting coverage for all the separate modules)

[protolambda]

Stuck on a go issue:

../../../go/pkg/mod/github.com/btcsuite/btcd@v0.22.0-beta/chaincfg/genesis.go:10:2: ambiguous import: found package github.com/btcsuite/btcd/chaincfg/chainhash in multiple modules:
        github.com/btcsuite/btcd v0.22.0-beta (/home/proto/go/pkg/mod/github.com/btcsuite/btcd@v0.22.0-beta/chaincfg/chainhash)
        github.com/btcsuite/btcd/chaincfg/chainhash v1.0.0 (/home/proto/go/pkg/mod/github.com/btcsuite/btcd/chaincfg/chainhash@v1.0.0)

Which I think causes a build issue which the linting tooling can't handle (but go itself can somehow...):

golangci-lint run -E asciicheck,goimports,misspell --fix
WARN [runner] Can't run linter goanalysis_metalinter: buildir: failed to load package : could not load export data: no export data for "github.com/btcsuite/btcd/chaincfg/chainhash" 
ERRO Running error: 1 error occurred:
        * can't run linter goanalysis_metalinter: buildir: failed to load package : could not load export data: no export data for "github.com/btcsuite/btcd/chaincfg/chainhash"

[protolambda]

See btcsuite/btcd#1839
Fixed by upgrading to github.com/btcsuite/btcd v0.22.1

[trianglesphere]

In general this looks good. The big thing I'd like is to set a cutoff date for which PRs go in first and which PRs should be built on top of this branch (to try to limit merge / rebase conflicts).

[protolambda]

Rebased to resolve merge conflicts. Edit: and force-pushed again to fix the opnode > op-node change in the new fuzz makefile target.

[protolambda]

Deleted code cov comment since it wasn't updating anymore. Coverage is stable, no changes to actual code, and we merge all the coverage reports of different modules by pushing them all to codecov. You can find the coverage diff here: https://app.codecov.io/gh/ethereum-optimism/optimistic-specs/compare/main...go-work (some renames, basically 0 coverage difference)

[trianglesphere]

I left some small comments around the makefile.
Also FYI - the integration test didn't pass locally, but I think that is probably the test being flaky b/c it passed in CI.

[trianglesphere]

Not here, but in submodules, we very likely want --recursive. Maybe this changes with the new contracts stuff, but foundry-lib seemed to need it.

[trianglesphere]

The clean command doesn't do anything b/c we leave binaries in the subdirectory

[trianglesphere]

Not this, but above COMPOSEFLAGS /ITESTS_L2_HOST doesn't seem to be referenced anywhere. (

[protolambda]

I can help fix this merge conflict with a rebase in 4 hours from now. Basically need to rebase on main, drop all the old file moves + import rewrites, and redo those, then fix up some go imports, while preserving the docker and ci config fixes.