Current supply on coingecko (17,634,537) and CMP (17,634,537) was not updated for a some time, because well, no any working blockchain explorer.
So I calculated it for the block 212000 (we at ~2129300 now) and added this info to my table.
about 17 874 521 GAP at the block 2120000.

Finally I was able to launch Qt wallet on linux. Previously I used
gapcoin-0.16.3-aarch64-linux-gnu.tar.gz, because I thought it's the only release for desktop linux,
and gapcoind and gapcoin-cli worked just fine, but not Qt.
But now I had found
gapcoin-0.16.3-x86_64-linux-gnu.tar.gz
Now not only I have GUI wallet, but also the same python script calling the daemon works 4-5 times faster it seems
You can find releases here:
https://github.com/gapcoin-project/gapcoin-core/releases
And I have no slightest idea how to build it by myself.

It seems sometimes we have 3 nodes on satoshi9.2 connected.
I can connect to them from Windows machine but not from Linux machine.
But they all disappeared today from my nodes even from Windows machine.
Maybe it has something to do with me adding port=8334 to gapcoin.conf, but maybe not and it's just a coincidence.
Because I could not connect to them from Linux machine at all, I forwarded a port and connected my linux desktop node to my windows node.
I guess now anybody can try to connect to it too (not sure if it will work):
178.141.100.234:8334
Main alive node de-facto is still:
88.66.203.169

About Nvidia miner: it actually works fine, I just forgot to add -a nvidia to .sh
It seems I can launch multiple instances and they do different work, but I'm not sure about it.
If it's indeed so, then from rtx 3060 with decent multi-core CPU we can get at least 8M primes per second, probably even 10-12M and more.
But it seems it's not possible to mine any shift but 32 on GPU miner. -f parameter is ignored. If anybody know how to change prime shift in
GPU miner, please share this knowledge in this thread. Even only working 16 or 64 shift will make mining much more fun.
On 32 shift we have the world record aka the biggest merit in the blockchain, no hope to overdid that.

I tried to build miner from source, CPU version no problem, it's compiled fine and works fine afterwards, but no luck with GPU version, probably because I tried to build CPU branch with GPU settings instead of GPU branch
(I've built some binary file for GPU eventually, but it's not working properly and file is different from released miner). Maybe I will try my luck with GPU branch later.

I think the only way for mining to survive is to somehow get some support from hobbyists, who will mine mostly just for sake of it. Even at 10 sat per GAP,
current daily profit from all mined coins is about $2 (0.6GAP*24b/hx24hx10sat). I can buy some from time to time, but it's basically a donation, I have no intention to sell them and I have no much hope that it will be ever possible to sell them later for profit (or even buy some goods, lol, it's a cryptocurrency after all).

Another thing is halvings. They are happen very fast. Honestly I have no idea how even Bitcoin itself will survive them. I prefer the Monero way with constant tail emission. Of course we have no coordination (and probably no skill to build a new wallet) to hard fork GAP for a constant 0.6 (current) or at least at 0.3 (next halving) per block. Primecoin way to go regarding block rewards is also much better in my opinion.

About "death of the coin". Question is: do we want such project to exist - not in a sense of cryptocurrency, but as a math distributed computing project? I think the answer is yes, so why to launch GAP-2 when we already have it? As long as somebody mine it, as long as coin is more or less alive, we have the blockchain with all results and also a chance to improve them. If anything, we need somebody from Primecoin or Riecoin team to be interested to support the project. Maybe some old whale can send somebody who can have some interest in GAP some tens/hundreds of thousands of GAP, so such hypothetical dev will have some incentive for price to rise.

Maybe I will try to make the rich list next, but of course, it is much harder task than shift/number of blocks/merit list and calculation of current supply, so I'm not sure that I'm capable of doing this.