fix: avoid unnecessarily splitting expressions with multiplication terms with a shared term

[TomAFrench]

Description

Problem*

Resolves

Summary*

This PR addresses an issue where we were unnecessarily splitting an expression based on an example which Zac found while working on noir-edwards. We were being overly restrictive and only accepting the case where both witnesses in the multiplication are shared with other terms

Additional Context

Documentation*

Check one:

• [x] No documentation needed.
• [ ] Documentation included in this PR.
• [ ] [For Experimental Features] Documentation to be submitted in a separate PR.

PR Checklist*

• [x] I have tested the changes locally.
• [x] I have formatted the changes with Prettier and/or cargo fmt on default settings.

[github-actions[bot]]

Changes to circuit sizes

Generated at commit: 0280a9c70681f8443729e777f6088ad21acd60b5, compared to commit: 3ef36458fef36b2a2f6cf99b35a43339f3721b27

🧾 Summary (10% most significant diffs)

Program	ACIR opcodes (+/-)	%	Circuit size (+/-)	%
hashmap	-4 ✅	-0.00%	-4 ✅	-0.00%
regression	-3 ✅	-0.86%	-3 ✅	-0.08%

---

Full diff report 👇

| Program | ACIR opcodes (+/-) | % | Circuit size (+/-) | % | |:-|-:|-:|-:|-:| | **hashmap** | 209,911 (-4) | **-0.00%** | 397,324 (-4) | **-0.00%** | | **regression** | 347 (-3) | **-0.86%** | 3,910 (-3) | **-0.08%** |