This picture says it all. In it, you can see the route a payment is taking across the Lightning Network. There are a _very_ large number of possible paths payments can take, and every node is close to every other node. Furthermore, there are privacy benefits to this arrangement.
Read about the possible privacy implications of Lightning here medium.com/@scidexer/incr… There is no argument against software like this, which is not only proven to work, but which is under withering scrutiny because money is at stake. On chain scaling looks absurd next to this.
Anyone who is really interested in Bitcoin's fast adoption and global domination is working with Bitcoin as the infrastructure. No other path is logical, safer, more scalable or beneficial. Logically if your aims are aligned with Bitcoin's goals, you will only work on Bitcoin.
There is a huge amount of work across all fields of software development that needs to be done on Bitcoin. Outsourcing the hardest work to the people best suited to do it, leaves you free to develop services and software that will serve billions. This makes perfect sense.
Building the services that will serve billions of people is not trivial, and requires disciplines across many different areas. It will have influence over the entire user space ecosystem, and touch everyone. Fighting over who runs the engine of the ecosystem is totally illogical.
And this is the problem; the people who think that the Bitcoin reference client is key are not computer literate or motivated by a desire to help others, and they also can't identify new problems that need to be solved. They can only see things in terms of what's happened before.
Bcash is not about innovation. It is a "Bitcoin Cult" in the same way that the famous islanders formed "Cargo Cults". You will never get an innovation like Lightning from bcash, nor will you get research and development that will cause their altcoin to beat Bitcoin. Good!
The real question is why are you asking me on Twitter? All the documentation to do with this is online, and you should go and read it yourself. It's important to talk about these things as a principle first; scaling by increasing the block size will ruin Bitcoin, so rule it out.
Second, you've just posed a Straw Man. What if all channels have enough capacity to send the smallest amount that anyone wants to send? Then everyone in that class of user is satisfied. With hundreds of thousands of nodes running simultaneously, everyone will be catered for.
Asking "gotcha!" / Straw Man questions is what bcashers do. It adds nothing, helps no one and creates no new software. If you're smart enough to ask the right questions, you're smart enough to write software for your own scaling solutions. If not, you write against existing APIs.