[BUG]: `plaintext_value` → `value` rename emits spurious update diff on every existing *OrganizationSecret

### Module

`github_actions_organization_secret`, `github_dependabot_organization_secret`

### Version

v6.12.0 (introduced by #3225)

### What happened

PR #3225 renamed `plaintext_value` → `value` (and `encrypted_value` → `value_encrypted`) on the secret resources and deprecated the old property names, but did not add a `DiffSuppressFunc` to smooth the migration. As a result, every existing resource that was created with `plaintext_value` in state now shows an `~ update` diff after the user renames the argument in their config — even though the secret's actual value is unchanged.

For our Pulumi wrapper (`pulumi-github`, bridged from this provider), we see 174 `~ update [diff: +value-plaintextValue]` entries across 11 stacks, one per `*OrganizationSecret` resource, after switching the single source-code call site from `plaintext_value=...` to `value=...`. Each update issues a GitHub API call that writes the same plaintext secret value back. Idempotent on the GitHub side, but noisy:

- 174 unnecessary writes to sensitive resources.
- 174 audit-log entries that look like real secret changes.
- State-side churn on what should be a pure property rename.

### What was expected

A `DiffSuppressFunc` on the deprecated `plaintext_value` (and `encrypted_value`) properties so that when the new `value` (or `value_encrypted`) contains the same data, no diff is shown — mirroring the pattern already established for etag properties in #2840.

### Reproduction

1. Create a `github_actions_organization_secret` (or `github_dependabot_organization_secret`) with `plaintext_value = "foo"` under provider v6.11.x.
2. Upgrade to v6.12.0+.
3. Rename the argument to `value = "foo"` (no value change).
4. Run `terraform plan` (or `pulumi preview`) — observe an `update` entry with a diff of the form `+value-plaintextValue` (or `+value-plaintext_value` in TF syntax).

### Suggested fix

Add `DiffSuppressFunc` to the deprecated `plaintext_value` / `encrypted_value` schema entries so they diff-suppress against their replacements when the underlying value is unchanged. The etag handling in #2840 is a good template.

Happy to open a PR if that helps — flag if you'd accept one.