Lightning removes the practical capacity ceiling

Routed payment channels remove Bitcoin's practical capacity ceiling: throughput becomes orders of magnitude greater with no practical upper bound, settlement drops from minutes to milliseconds, and payments can be smaller than the on-chain dust limit.

For (strongest version as argued): The ceiling is a property of the base layer, not of Bitcoin. Two parties exchange signed commitment transactions that never touch the blockchain and that no one else ever sees; only the funding transaction must be mined to open the channel. With relative timelocks (CHECKSEQUENCEVERIFY) a channel can stay open indefinitely, and the notes report prototype implementations supporting more than 281 trillion state transitions in a single channel. Because value moves without committing a transaction to a block, HTLCs clear in milliseconds rather than minutes, at satoshi-level granularity — at least as small as the dust limit and perhaps smaller. Nothing in that construction has a throughput bound that scales with block space.

Against (strongest version as argued): "No practical upper bound" conflates the capacity of a channel with the capacity of a network. A channel's state transitions may be effectively unlimited, but the value that can move through it is bounded by what was locked in the funding transaction, and the value that can move along a route is bounded by the smallest balance on it. That is a real ceiling, differently placed. The network also inherits the base layer at exactly the two moments that matter for growth: every channel must be opened on-chain and eventually closed on-chain, so onboarding is rate-limited by block space, and a period of high on-chain fees raises the cost of both.

And the theoretical ceiling is not the binding constraint anyway — adoption is. A second, independent account (The Blocksize War) records that in 2015 Expedia, Overstock, TigerDirect, Newegg, Dell, Rakuten and Microsoft all accepted on-chain bitcoin, and concludes that on this narrow point "large blockers have been shown to be mostly correct": that on-chain merchant adoption at the end of 2015 appears to have exceeded lightning adoption as of that book's writing. Many of those merchants later stopped, so this is not an argument that on-chain payments won — it is evidence that a capacity ceiling being raised in principle did not, over roughly a decade, translate into more payments in practice. The same account lists the practical obstacles: channel liquidity, the requirement that the receiver be online, credit risk on funds locked in channels, and the ongoing burden of preventing theft from one's own channels.

Nuance: The claim bundles four properties that succeed or fail independently: throughput, latency, granularity, and capacity. Latency and granularity follow fairly directly from the mechanism described. Throughput and total capacity depend on network topology, liquidity distribution and routing success — none of which the mechanism source models. A debate that establishes "settlement is fast and payments can be tiny" has not established "the scaling problem is solved," though the two are commonly presented as one point. Note also that the two sources here are arguing about different things and are not straightforwardly in conflict: one describes what the protocol makes possible, the other reports what was adopted. A careful debater should not treat the adoption datum as refuting the mechanism, nor the mechanism as answering the adoption record.

Sourcing: two independent owner-research files by different authors, reaching opposite practical conclusions — which is why this sits at CONTESTED with medium rather than low confidence. Neither is a primary source, so INGEST_PLAYBOOK.md still caps the status at CONTESTED. Both describe an era rather than the present: neither file states an edition or publication date, so every figure here is a snapshot of its book's writing, not a current network measurement. The adoption comparison in particular is undated and should be re-verified before it is used on stage.

Common misstatements: "Lightning gives Bitcoin unlimited transactions" (the source's claim is about a channel's state transitions and the absence of a block-space bound, not the absence of liquidity bounds). "Once you're on Lightning you never touch the chain" (opening and closing are on-chain by construction). "Lightning already handles the volume of a card network" (a throughput claim requiring measurement that neither this source nor this corpus supplies — see lightning-network-basics for what is actually established).