Policy: Penalize effective fee for sub-dust outputs

[kwsantiago]

Closes #146

Reduces the effective fee by the dust threshold for each sub-dust output, creating an economic disincentive for dust creation without outright rejecting transactions.

• New -subdustfeepenalty option (default: enabled, disabled by -acceptlargefees)
• GUI toggle in Options
• Functional test covering penalty, proportional penalty, stacking, OP_RETURN exemption, and disable flag

[luke-jr]

How would this work, since dust is just rejected entirely still?

[kwsantiago]

The penalty applies to sub-dust outputs already tolerated by other policy settings (ephemeral dust, acceptnonstdtxn, etc.). Per #146, it targets tolerated sub-dust outputs rather than overriding the dust rejection. OP_RETURN is unaffected since GetDustThreshold returns 0 for unspendable outputs.

[luke-jr]

Should "test: fix subdust fee penalty test for corepolicy min relay fee" be squashed?

Note for followup later: This doesn't allow for at least two useful configurations:

1. Reject dust below fixed dust floor, but penalise based on smart dust logic.
2. Allow all dust, but penalise based on dust configuration (fixed or smart).

[kwsantiago]

Squashed the test fix, moved SoftSetArg to the default-toggling commit, and removed subdustfeepenalty from the rejectunknownscripts toggle.