Back then it was less characters and far less settlements: the script is now much larger plus some other monitors to handle trait checks etc.
The monitors themselves are fairly harmless so far as turn end processing goes (I believe) except perhaps CharacterSelected. I have purposely avoided CharacterTurnStart/End - the biggest turn end killer so I've heard, which stands to reason given the number of characters on a map.
The script is about 80,000 lines but 99.3% of it is all in one monitor - the main, process-heavy one - and that only fires when the player begins a switch. Never triggered during turn end.
But all of this script still needs to be read and kept in memory by the game. So the question is how much does this 'chewing up memory' affect turn ends? I haven't tested it but I did test this kind of thing when building the Palantiri which is about 50,000 lines and similar design (all the real processing in one huge, player-triggered-only monitor). I don't have the numbers on hand (buried in a PM somewhere) but it was something like half a second added to the total turn end time. And some of that would have been caused by Palantiri trait/anc/other triggers, not the hugeness alone.
I'm not sure how much has changed that would affect it. Character names, unit names, different bodyguard options including new units, hero abilities (these are reapplied by the script), settlement internal names and locations, new settlements, ...
When I tried running the tool recently it crashed, due to a peculiar entry in 3.1's settlement/region name file: it has {Gondor}Gondor at the top. I need to adjust the tool to allow for this kind of thing; a redesign of how to read the files.
So thanks for the offer Baron but it would be a heroic effort just to figure what needs to change, even before getting to making the changes. The tool makes very short work of it, so it makes sense to let the tool do it. I 'just' need to fix it.
I have to say though that response to the release was poor (though the response to the idea was good ). Not many downloads and absolutely no feedback, apart from killersmurf. Hence being low on my priority list. I spent a ridiculous amount of time making it; if interest in a 3.1 version is low (i.e. few requests for it) then my interest in more time spent is low too. I'd much rather spend the time on new things.