Revert potentially dangerous optimization

[EndymionJkb]

Description

As Juani pointed out, caching the bufferWrappedSurplus might be dangerous, given that there are external calls between when it's initialized and the last time it's used.

I tried a few variants that preserved some of it (e.g., just reloading as necessary), but as the point of this was to save bytecode and get the Vault to fit, simply reverting that change turned out to be best. (The variants were actually slightly more bytecode.)

Luckily, due to the other changes, it still fits with this change reverted.

Type of change

• [X?] Bug fix
• [ ] New feature
• [ ] Breaking change
• [ ] Dependency changes
• [ ] Code refactor / cleanup
• [ ] Optimization: [ ] gas / [ ] bytecode
• [ ] Documentation or wording changes
• [ ] Other

Checklist:

• [X] The diff is legible and has no extraneous changes
• [ ] Complex code has been commented, including external interfaces
• [ ] Tests have 100% code coverage
• [X] The base branch is either main, or there's a description of how to merge

Issue Resolution

[EndymionJkb]

Looks good now!

One final ask: can we make the code symmetrical? Either option works (make the surplus variable as local as possible as wrap, or declare above and reuse as with unwrap).

Keeping the exact same structure helps readers follow through better. Same that happens with add / remove.

Sorry for being so annoying :D

NP; will do