feat(fill,docs,fw): improve `fill`'s test dev-ex

[danceratopz]

EEST test developer experience should be improved, especially to help new testers. There's definitely a few low hanging fruit that could really help test writers.

Plan

• Get feedback from newer test developers.

Framework

• Ensure EEST types have good (concatenated) string representations to help debug test fails (via __repr__ or even __rich__? rich offers more customized formatting and can even display table representations, etc.).
• Improve tracing output, cf all: Trace types #474

Fill

• Introduce standard terminology for "fill test failures" - be consistent in docs/docstrings/error messages.
• Detect OOG without having to enable tracing and manually check the traces. At least point a test implementer to enable tracing if we don't enable it by default.
• Consider printing more mismatching storage keys if more than one differs (instead of only failing on the first detected mismatch), perhaps only with increased verbosity (-v or -vv).

Tooling

• Add trace diff tooling: cli tools, Claude backslash command, ...

Docs

• Improve documentation on how to set gas limits (to avoid failures).
• Add best practices and troubleshooting section.
• Add a section on how to debug a failing test using fill --pdb and inspecting objects, respectively using rich.print() in pdb.

[LouisTsai-Csie]

Consider printing more mismatching storage keys if more than one differs (instead of only failing on the first detected mismatch), perhaps only with increased verbosity (-v or -vv).

Suggested format:

<account>
<slot a> <expected> -> <actual>
<slot b> <expected> -> <actual>
...
<account>
<slot a> <expected> -> <actual>
<slot b> <expected> -> <actual>
...