I think it is more natural because it tackles things like exception handling in a new novel direction: Zero-overhead deterministic exceptions, among other things. He picks up all the things that clearly have great potential (and some have been proven to be very good in other languages) and is making a frontend for them.
Carbon, on the other hand, is just Google code standard as a frontend. No exceptions, weird naming schemes etc. It's like a coding guideline made language. The guys behind it clearly know performance though, so I am not saying it won't be that. But will it be a great language? I don't think it can compete with what Herb Sutter is trying to achieve. I am of course willing to be proven wrong.
As I see it, Carbon have in its explicit goals to fix both the syntax issue and the major ABI retrocompatibility issue, while herb approach is to just deal with the syntax: Sure it is better than nothing, but I'm willing to bet that this will not convince most of people that saw their idea being killed by the committee on the ABI altar.
Carbon, on the other hand, is just Google code standard as a frontend. No exceptions, weird naming schemes etc. It's like a coding guideline made language. The guys behind it clearly know performance though, so I am not saying it won't be that. But will it be a great language? I don't think it can compete with what Herb Sutter is trying to achieve. I am of course willing to be proven wrong.
See: https://www.open-std.org/jtc1/sc22/wg21/docs/papers/2019/p07...