Re: Is ICANN even needed ?
This line of argument is so wrong that I don't know how to begin.
1. IPv4 - it's true that this is now effectively entirely in the hands of the regional Internet registries (RIRs), or soon will be. But we still need a central registry of the space allocated to the RIRs and of the legacy allocations to non-RIR customers.
2. IPv6 - addresses are not allocated to countries, but to providers and major (provider-independent) sites, again by the RIRs. But we stiil need a central registry of the space allocated to the RIRs and the vast unallocated space.
3. "Assign the country designators (e.g .uk for the United Kingdom) to their governments to assign." Didn't you read the stress test scenario about Ukraine?
4. "Requests for new non-country designators could be handled by a small subgroup of the ITU." I think you will find that both ICANN and the vast majority of its supporters and detractors all agree that the worst possible solution would be the ITU.
5. You have ignored the fact that most of IANA's technical work, currently performed by the IANA team at ICANN, is the registration of protocol parameters.
BTW I think the set of stress test scenarios is really good work.