The purpose of a block reward is both to secure and to distribute. In the previous discussions I think it became clear enough why a linear emission is fair for current and future generations to have the same amount of grin being emitted they can buy, see this thread:
So let’s focus on the aspect of security for this research.
I think there is one major omission in this research, you only look at the absolute security of security. IMO it makes more sense to look at relative security
Relative security would be:
dollar_value_mined_coin / dollar_value_market_cap
In which case grin outperforms all coins after their first halving. I mean if there is little to be gained by an attack due to low value of a coin, proportionally, less security is needed assuming an attacker would be in it for the money only. An alternative method of calculating would be:
daily_volume_value /daily_emission_value
Also using this alternative way of calculating Grins relative security would outperform these other coins after their first halving.
Other factors that should not be ignored IMO are ASIC versus GPU and CPU minability. CPU and GPU mining provide much more opportunity for an attack since the attacker does not need to invest in hardware specific for a coin.