Skepticism is good, it helps weed out ideas that stink,
Definitely my design philosophy. Trial by fire!
merely a peek at what for the human player are interface accessible settings (sliders, research, civics, troop numbers and placement, foreign deals etc.)
It doesn't seem unreasonable, I just think its important to make sure that its actually valuable info.
I guess my thoughts go this way: you shouldn't have to expend any effort/resources to gain information that:
a) Would be patently obvious IRL (eg: civics, you don't need to spy on them to know if they're a monarchy or a democracy, or if they support free trade or mercantilism, you can see that by looking out the window). Similarly, what tech the enemy has should be pretty apparent; its not that hard to figure out if they have Railroad yet, or gunpowder.
OR
b) Wouldn't affect your decisions/actions once you know them. I can see myself making different decisions if I observe that they have a huge army, but that's about it. I'm not going to change my own actions at all if I know they're 60 science/40 gold vs knowing they're 70 gold vs 30 science, particularly when they can easily change that from one turn to the next.
I would defininitely say that if there was a military slider such as you suggested, it would have to be something that the AI would respond to, in the same way they the Victoria AI responded to you mobilizing your reserves.
I dislike the idea of the AI using "sleight of hand". It becomes very difficult to distinguish between an AI that is being "tricksy" and an AI that is just bad/confused.
I agree that the espionage system performed poorly, the AI was weak with it, it was MM intensive and most of the missions were blah. We're trying to work on a more interesting design for the Dune Wars total conversion mod, where the missions available are either mroe powerful or cheaper, and the mission accessibility varies across faction (and each faction is limited to a handful of missions).
Certainly the removal of access to diplo modifiers (I believe that was confirmed) suggests we will need other clues to determine each AI's likely behaviour.
You may have seen my reaction to that in another thread...

I think its just poor design to make the AI opaque, because it makes it too hard for the human player to actually figure out an intelligent diplomatic strategy.
Local sliders would allow for example some 'center of learning' cities to run at a larger deficit and other 'gold-producers' to run at a large surplus, but balancing out over the entire empire. It would allow for heavier city specialization but still preserve the global rule that doesn't allow accumulating a national debt.
Ah, I see. I think I mentioned localizing sliders earlier in this thread.
As I mentioned, I think the problem is that it promotes *too* much specialization. You end up setting every city to either 100% science or 100% gold, and only building the appropriate booster builidngs in each. And I don't think that's what we want.