The observations made on the code are correct logically but not universally. Here is my counter:
Loops can be expensive in solidity however they depend on a large number of factors. A loop that write a storage variable in every cycle is expensive or a loop that sends out tokens or ethers in every cycle. In-memory loops are not that expensive - especially given the loop in the function stops when it finds the first memory candidate to alter(slash).
If you look at SHUF code - there are loops (while limited to 512 elements) - they essentially run sorting algo - which means when it finds a new address with a +ve token value, it not only has to run a loop but also alter the order of multiple storage elements to keep the rank list intact (bubble up/ down).
Now realistically it's highly unlikely that ALIVE will have users that run into tens of thousands. I know I am FUD-ing my own project but it's the truth. SHUF, on version three, and after a generous and wide airdrop, has 3000 holders approximately. I am not trying to use SHUF as a shield to justify bad code or trying to poke holes there. I have a lot of respect for the project because I learnt from it. What I am saying is that a code can theoretically break doesn't mean it will practically break.
ALIVE at this point has 147 addresses with 93% of airdrop already done. Right now most of the gas in the slashing function is being used for writing the storage elements. I don't see any cause of concern due to the loops in the near future. And if I end up being incorrect - then it'll be learning.
Having said all of that I have already mentioned that the objective of this project is not to be an investment vehicle.
Last but not the least, I am using this to learn for a large DAPP i am building which will be a full scale game. I am developing that game on my own time, with shoestring budget and will be free for all to play forever. There would be tokens in that but with simpler logic and the burden of the game will be carried on by other non ERC20 functions. It's difficult to get a large group of folks to test a demo and so a gamified token was the best option for me. ALIVE is still an independent project but all the feedback is helping me understand dynamics when a larger group is involved.
After reading your conversation with Agusx1211 on Discord, I'll like to know if a new token will be created or not. Thanks