impl compact<()> for MaxEncodedLen

[pgherveou]

Third time's the charm
Adding impl compact<()> for MaxEncodedLen to get polkadot-sdk to compile with this branch

see paritytech/polkadot-sdk#1424

[pgherveou]

I think that one is necessary for all tests to pass on polkadot-sdk, will run CI against it to make sure this time

[pgherveou]

Tests seems to pass on polkadot-sdk with this branch
paritytech/polkadot-sdk#1424

It's just failing these 3, but I believe that might be related to network access to github

@paritytech-ci
continuous-integration/gitlab-build-push-image-malus — {"build_allow_failure":false}
[Details](https://gitlab.parity.io/parity/mirrors/polkadot-sdk/builds/3618748)
@paritytech-ci
continuous-integration/gitlab-build-push-image-polkadot-debug — {"build_allow_failure":false}
[Details](https://gitlab.parity.io/parity/mirrors/polkadot-sdk/builds/3618739)
@paritytech-ci
continuous-integration/gitlab-test-frame-ui — {"build_allow_failure":false}
Required
[Details](https://gitlab.parity.io/parity/mirrors/polkadot-sdk/builds/3617838)
@paritytech-ci
continuous-integration/gitlab-test-linux-stable 3/3 — {"build_allow_failure":false}

[KiChjang]

It would also be nice to try and figure out where we use Compact<()>, because at first sight, it looks very weird to see that we're trying to compactify a ZST.