It's the transition of power that's the problem. Like, a real bad problem. Like raze half your country and set you back 50-100 years problem.
Roman dictators weren't like modern autocrats. They were officials, appointed for a limited time (usually no more than 6 months) to address some specific emergency. They weren't crooks who staged coups. And it doesn't seem to have been a desirable appointment to hold; dictators often stood down before the time-limit.
I think the GP's point was that even with "good" dictators, the problems come up when succession needs to happen, which is why "dictatorships" don't scale.
Democracy has a built-in mechanism to stop any bad thing from happening in a change of government, while ensuring that change happens often (at least in a good one).
A ruler with absolute political power.
> ...modern version, of which there are many terrible examples before power transition becomes an issue?
This actually reinforces my point. I never denied the existence of bad dictators (or even that the majority of dictators are bad, I think that's quite likely). This would make the power transition for a country that happened to get a good one significantly more dangerous.
That, and the lack of freedom, and the life you're subjected to when the dictator doesn't like you.
It does have a massive mall attached to the public area though (Jewel).
https://travel.stackexchange.com/questions/34818/why-is-secu...
It is helpful in the sense that bottlenecks don't happen earlier on, but I'm not sure it's that great of an approach.