I address all opponents of Layer2 mechanisms.Bitcoin transaction protocol is very similar to roads. For my comparison, I will make the following assumptions:
individual, passenger/driver = 1-to-1 transactions
carpooling = 1-to-many transactions
fee in sat/byte=potential car speed or speed-denominated lanes (40 mph, 55 mph, 70 mph, 80 mph)
Like the road's main function is to allow the transportation of people and goods, Bitcoin's function is to transfer value and messages from individual to individual.
Let's assume economic growth, i.e. increased demand - the transportation intensity increases. The individual faces 3 choices and infrastructure has to cope with:
1. drive faster and more often - it's more expensive and has environmental costs; road maintenance cost rises due to increased usage;
2. drive larger vehicles - similar; similar and new lanes are required=additional CapEx;
3. pool with other participants - marginal costs decrease, improved efficiency; reduced CapEx compared to p. 1-2.
After some period, the transportation flow is so dense, that p. 3 reached its peak efficiency. Cars collide, intersections are overloaded, people and cargo reach their destination with delay. Whenever a scaling issue appears it can be solved:
a) Extensively - building new lanes and enlarging intersections. This is the preferred argument of dumb people, that don't care about efficiency - "I need space for my wide body car, even though I am driving alone" (rednecks with Dodge RAM 3500, hello!). In blockchain it equals to larger blocks.
a1) Extensively - increasing speed limit and building new lanes. With increased speed limit and lack of discrete infrastructure, more accidents will happen and the energy consumption will also increase (drag is proportional to the speed square, hence poor energy efficiency). Increasing speed limit=faster block time, more lanes=bigger blocks.
b) Intensively - prioritizing lanes, adding fast lanes, roundabouts and multilayer intersections and the most important, introducing public transport with dedicated infrastructure: trains, subways, Bus Rapid Transit, even pneumatic tube transport or HyperLoop. All these improvements have marginal benefits, as unitary cost of moving 1 person or payload decreases and security increases.
In blockchain, I view similar concepts in:
Lightning Network - value transfer between related parties, with the clearing of the final output;
SegWit - similar to buses and trains that transport a group of people, SegWit groups transactions together and deliver them faster, cheaper;
Child chains and off-chains - private roads or transportation means within a closed group;
Privacy features (Tumblebit, Sphinx) - the random taxi you take to the motel
Before debating me, please check the traffic flow simulation - extensive vs intensive scalability:https://www.youtube.com/watch?v=yITr127KZtQ
