The one failed file is nothing to worry about. Terminator is adding some new options to the updater package and one of the support files is missing. I'm sure it'll be taken care of shortly.
The LandRatio is a dynamic value but I probably should explain it better.
As you may be aware, every unit in the game has a threat value, and one of the reasons that LOUD requires it's own unit database, and also why we adapt any mods we use, is that we have a series of formula that are used to give every unit in LOUD a threat value that is relative to what it is - and what it does. In this way - when LOUD is looking around the map, he can make much more informed decisions about where he can go and fight and where he cant and if it's worth it do so or not. LOUD will not be a mindless automaton and just throw himself away just because he can. The 'kitchen sink' attack is simply a 'pressure valve' that will flush him out, mostly to minimize traffic issues, but often to allow him to replace them with bigger ticket items.
The LandRatio is an overall average of enemy LAND threat versus his own overall LAND threat. Now - having said that - it's not perfect - it's only going to be as accurate as his intelligence gathering - and LOUD knows that - so he has other similar metrics that tell him if he's getting his intel reliably or not - and that too will impact his evaluation of threat.
At this point - as we tune LOUD in more specific ways - we are thinking that we may be able to reintroduce a feature that was laid out in the game, but never used. AI Personalities.
AI Personalities would be a small table of 'adjustments' to those trigger points that would allow one AI to be more aggressive while another would be less. Without the human player knowing what he's getting when the game starts.
Going back to your comment about scalability - there are a number of interrelated issues and their not all about the AI and his cheat level. More often than not, the map, the amount of mass, unit cap limits and the number of players are more a defining matter than the AI. That's why the LOUD team spends as much time going over maps as we do, mostly to find situations that put the players into team-oriented tactical situations that require a collective solution. Generally, we prefer the larger maps as we see them as providing an opportunity to push the player out onto the map - and give all that real estate some value - but also make the 'shoulder to shoulder' team defense situation much harder to come by.
Again, thanks for your comments. They all help us refine LOUD each and every day.