CVSSv4 Reference Score Discrepancies

While qualifying my own CVSSv4 implementation, I found that 159 entries in the retained [`reference-scores`](https://github.com/FIRSTdotorg/cvss-resources/tree/48f85d84a036de9c610668f9496c12b5040a9ae3/vectorFiles/reference-scores) corpus disagree with the later RedHat calculator revision.

Observed discrepancies:
- 159 affected corpus entries
- 157 unique vectors
- every score changes by exactly +0.1
- no resulting severity changes

The differences appear to follow the floating-point adjustment introduced by [`RedHatProductSecurity/cvss-v4-calculator@7faaaa5`](https://github.com/RedHatProductSecurity/cvss-v4-calculator/commit/7faaaa5662d7c94e160c1193c96a970ced773301). I have retained an [executable generator](https://github.com/cticommons/cvss/tree/v1.1.1/differential/cmd/cvss40-corrections) which verifies the pinned calculator and corpus by exact length and SHA-256 before reproducing the [157 corrections](https://github.com/cticommons/cvss/blob/v1.1.1/testdata/first/v40-rounding-corrections.json)

Is reference-scores expected to be updated for the later rounding behaviour or is the older output intentional?

Start by reviewing the pinned reference-scores corpus and the linked RedHat calculator revision, then run the retained differential/cmd/cvss40-corrections generator. Compare its verified 157 corrections with testdata/first/v40-rounding-corrections.json; the issue is resolved when the expected corpus behavior and whether an update is needed are established.

Written by the indexing model from the issue text.