Thanks for the suggestions and I apologize for lack of communication on this important issue. The main reason is that we are still struggling with what is the best move for the next step. We originally plan to use our slightly tweaked version of Cryptnoght R (and call it Cryptonight KV). However, that would require us to submit patches to various miners (e.g. XMRig, XMR-stak, T-rex, etc.). Since CN/R is a GPU/CPU algorithm, we need to implement support AMD, NVIDIA and CPU. It is quite a lot of work and uncertainty (not sure if the miners would accept our patches and they are moving to CPU-only RandomX). Then Kevacoin source code will need lots of changes. Unlike Monero source code, which is designed with hard-forks in mind, the Keva source code does not have support for hard-fork and we need to implement and test it. This is lot of work and testing. Honestly, we are secretly hoping that Nicehash's support for CN/R will go away, just like other older CN algorithms. We had a glimmer of hope this morning as the available hash rate of CN/R dropped to 1MHz (from 6 MHz) on the US market for a while.
However, we do agree that the CPU only is the future, and it fits our targeted developer demography, as these people normally have machines with powerful CPUs but only average GPUs. The recent development of AMD multi-core CPUs also makes CPU mining very promising. Our concern is that some of our early supporters are GPU miners and we don't want to disappoint them. But it is a fact that CPU mining is good for the long term for Kevacoin.
We will have a vote on the switch as suggested. Comments and suggestions are welcomed!
One thing that you can do, to not disappoint the early GPU miners is to premine the existing coins on the new blockchain and swap them for old coins so they won't loose the coins mined. Then they can continue to mine Keva like other if they want to stick to the project. This is how I would do it, as I think you can't simply switch to new mining algo, but rather launch a new blockchain.