feat: Refactor gas-optimized tx path for Ledger (refactor/gas-827-1)

[indocs]

The changes introduce a streamlined transaction encoding path and centralized gas accounting for Ledger-supported flows. Key updates include:

• Consolidation of the tx builder and signer interfaces to reduce redundant state reads and minimize stack depth.
• Updated encoding/decoding helpers to use inline assembly where safe, cutting several gas-heavy parameter extractions.
• A new optional gasEstimator pathway that leverages already-calculated deltas to avoid recalculating gas usage multiple times in cross-contract calls.
• Adjusted tests to cover the refactor’s edge cases, including bundled nonce handling and error propagation through the ledger signer.
• Migration notes: if you previously relied on the old tx path, ensure you migrate callers to the new builder API and update any mocks accordingly.

Gas trade-offs:
This refactor targets lower gas usage by avoiding repeated reads and by moving invariant computations out of hot paths. The trade-off is a slightly larger initial function footprint and increased complexity in the signer path, which may affect readability. When used in conjunction with the existing ledger optimization, total gas per tx should decrease on common paths, while some atypical flows may see marginal increases due to new inline checks. Overall, expected net gas savings should be observable across bulk-tx scenarios and higher-volume deployments.

[indocs]

Security nit: verify no reentrancy or unexpected external calls are introduced by the new modular path; confirm all external interactions are strictly via vetted interfaces and guard against untrusted call data.

[indocs]

Testing suggestion: add property-based tests or randomized input tests for the gas-optimized path to stress test under varying ledger inputs and ensure gas consumption stays within expected bounds across scenarios.

[indocs]

Developer-experience remark: consider centralizing common testing utilities (mock ledger, helpers for gas measurement) to reduce duplication and improve onboarding for new contributors.