External tests maintenance

[nikola-matic]

• Added reports/ directory to .gitignore, generated by running external tests.
• Changed hardhat/2115 and hardhat/2453 to point to hardhat/3365, as the former issues were closed in lieu of 3365, which is meant to encompass all such issues. I've also ping @fvictorio to see if hardhat/2051 falls under the same umbrella.
• Removed peg from ens.sh as the underlying issue was fixed.
• Removed hardhat peg from gnosis.sh as the underlying issue was fixed. This fixes a part of Remove workarounds for ethers 5.6.2+ and getStorageAt() from external tests #13051, but unfortunately the rest of the issues have not been solved, and the issue can thus not be closed.

[cameel]

By the way, this is a situation where separate commits would be perfectly appropriate. Each one of the points in the description is pretty much an independent change.

[nikola-matic]

By the way, this is a situation where separate commits would be perfectly appropriate. Each one of the points in the description is pretty much an independent change.

Separate commits or PRs? I theory I agree, but having 3 commits seems like overkill to me, and we're gonna squash them in the anyway, no?

[cameel]

Separate commits or PRs? I theory I agree, but having 3 commits seems like overkill to me, and we're gonna squash them in the anyway, no?

Commits. And no, we don't have to squash them. We only really squash review fixes into the original commits. Or we squash everything when the commits are so messy that they make the history harder to understand than just having a single commit. But overall all that we require is clean history, going straight to the point without any of the dead ends you went through when working on a solution, that is easy to understand when you look at it in git. Having each logical change in a separate, atomic commit with a clear message is fine and actually recommended.