chore: fail-open-multichain assets controller filter logic

[sahar-fehri]

Explanation

Restores the original fail-open behavior in MultichainAssetsController's Blockaid spam token filter, ensuring legitimate tokens aren't blocked due to API failures or missing data.

References

Checklist

• I've updated the test suite for new or updated code as appropriate
• I've updated documentation (JSDoc, Markdown, etc.) for new or updated code as appropriate
• I've communicated my changes to consumers by updating changelogs for packages I've changed
• I've introduced breaking changes in this PR and have prepared draft pull requests for clients and consumer packages to resolve them

---

Note

High Risk
Changes the spam/malicious token gating behavior so tokens may be added when the Blockaid bulk-scan call fails, which can reduce protection during outages and affects security-sensitive filtering logic.

Overview
Updates Blockaid-based filtering for newly detected token: CAIP assets to be fail-open: if a bulkScanTokens batch rejects, tokens in that batch are now allowed through instead of being dropped.

Adjusts unit tests to reflect the new behavior, including scenarios where bulkScanTokens throws or where only some scan batches fail, while still filtering tokens explicitly marked Malicious in successful scan responses.

^{Reviewed by Cursor Bugbot for commit 3be6a8c. Bugbot is set up for automated code reviews on this repo. Configure here.}