WITNESS DISCOUNT removal.
Keep in mind this is from uneducated non dev guy discussing with AI. 
Lightning Network – channel‑funding and closing txs are almost all P2WSH/P2WPKH, so their bytes now enjoy the 75 % witness rebate; killing the discount would roughly 4× the on‑chain cost of every open/close, splice, or anchor.
Taproot Assets / Taro – each asset issue or transfer commits a Taproot tweak or script revealed in the witness; its fee advantage disappears if witness bytes are repriced, so asset mints/ burns and Lightning‑routed asset swaps get four‑times dearer.
RGB – anchors bind a Merkle‑root to Bitcoin inside SegWit (key‑tweak or signature‑to‑contract). The scheme counts on 4 MB‑equiv block room; without the rebate epoch anchors become pricier and block room shrinks to 1 MB, forcing rarer batching or bigger fees.
Ark payment pools – mass exit or rebalancing dumps huge Merkle branches into a single witness‑heavy tx; the design’s fee math assumes discounted bytes. Remove the rebate and one emergency exit can exceed the old 1 MB cap or cost 4× more.
Bitcoin rollups (Citrea, Rollkit‑bitcoin‑da, Sovryn, …) – batch calldata/state diffs are kept in witness for the cheap 1 WU/byte pricing. Repricing would quadruple data‑availability costs and cut usable block DA bandwidth to one quarter.
Fedimint & Cashu – gateways peg BTC in/out through native‑SegWit bc1 outputs and rely on Lightning swaps; removing the discount would raise every peg‑in/peg‑out and liquidity‑rebalancing tx about 4×.
Lightning Network – channel‑funding and closing txs are almost all P2WSH/P2WPKH, so their bytes now enjoy the 75 % witness rebate; killing the discount would roughly 4× the on‑chain cost of every open/close, splice, or anchor.
Taproot Assets / Taro – each asset issue or transfer commits a Taproot tweak or script revealed in the witness; its fee advantage disappears if witness bytes are repriced, so asset mints/ burns and Lightning‑routed asset swaps get four‑times dearer.
RGB – anchors bind a Merkle‑root to Bitcoin inside SegWit (key‑tweak or signature‑to‑contract). The scheme counts on 4 MB‑equiv block room; without the rebate epoch anchors become pricier and block room shrinks to 1 MB, forcing rarer batching or bigger fees.
Ark payment pools – mass exit or rebalancing dumps huge Merkle branches into a single witness‑heavy tx; the design’s fee math assumes discounted bytes. Remove the rebate and one emergency exit can exceed the old 1 MB cap or cost 4× more.
Bitcoin rollups (Citrea, Rollkit‑bitcoin‑da, Sovryn, …) – batch calldata/state diffs are kept in witness for the cheap 1 WU/byte pricing. Repricing would quadruple data‑availability costs and cut usable block DA bandwidth to one quarter.
Fedimint & Cashu – gateways peg BTC in/out through native‑SegWit bc1 outputs and rely on Lightning swaps; removing the discount would raise every peg‑in/peg‑out and liquidity‑rebalancing tx about 4×.