First of all, please, do not be fooled by my low play time - I bought the game quite some time ago and have played it for hundreds of hours. And I am long standing fan of the series, starting SR2010, now playing SR1936.
To sum it up, my recommendation is positive, but, please, read more as there are few objections and e.g. Bjoern has quite a few valid points. I would not recommend this game to just anyone.
The game is very ambitious, wide in scope and quite complex with sharp learning curve. Of course you can assign some tasks to AI, but why let computer play my game? Especially as decisions of AI advisors are often quite poor. In general, game more or less delivers what it promises - choose any country and run it as you wish in global environment.
However, there is quite a feel of unfinisheness in some aeas. I think the scope of project was too wide to deliver within set budget and timeframe. Quite a few ideas are implemented only partially or got axed in full (NBC, environment, tourism and business centers... )
Now, to cover aspects of the game:
Research: there are several hundreds of technlogies and units in technology tree, but their effect is very often "research and forget" without any significant effect. Still, quite a few technologis do have very visible effect. As for the units, you can do well enough with lets say 7 units for most of the game (see warfare). Researching hundreds of unit models covered in game is marginally interesting. Tech level is a bit strange concept - it is possible to have tech level of year 1990 and run phaser hover tanks...
Diplomacy is quite strange (but I fail to recall any 4X game with really working diplomacy model, and I played many of them - Galciv probably got closest). E.g. casus belli is mostly based on active unit production capacity which gives way to many exploits (as AI is unable to control this aspect), but you will always face up "all computer coalition" in the end. However, I find diplomacy no better or worse than in any 4X game.
Economy is based (I believe) on Joseph Stiglitz theories. General idea is the more government spends, the better economy performs. Economy view aside, this leads to not very realistic in game strategies and exploits. It is very easy to start as - e.g. - Madagascar and triple GDP/c in a year. Moreover, there is very nasty bug related to price mechanism which I will cover later. I do feel economy to be somewhat ovrsimplified, but working (but mind you, I am economy geek :) ).
Warfare is acceptable. Building few hundreds of APC (preferably laser ones) and several naval transports is enough to conquer the world. Everything else is just nice to have. Long range AA and fighters helps limit APC loses against air powers, few dozens of stealth srategy bombers (B-2) are great to override entranched aeas. I found very limited use for tanks, and no use for AT, Recon, most naval units, range AA and most of the aircrafts... There is just no real need to build these types of units. AI does its job to provide resistence, but is by no means real oponent.
As for the nasty, game and save breaking bug I mentioned before... Using artifical production buildings (like synth fuel factory) does increase the final product price. This has positive feedback to output price of buildings which use his product as input (e.g. oil power plant), especially, if it the only input. If ANY country manages to create "closed circle" - like most of the oil produced in synth fuel plant, most of the electricity (used in synth fuel plant) produced in oil power plant - prices will get out of any reasonable range. GDP skyrockets and after several weeks treasury of this country will get out of range which game can handle. This destroys world market and after another few weeks, population gets out of control (not sure if by standard game mechanisms or the data just get overwritten). The game is not salveable in this state. There are workarounds (either conquer the country before it gets there - only handful of countires can get into this loop), or casually check the countries and if they start going this way, load one month old save in multiplayer mode and destroy these buildings. BTW, with careful handling this bug can be used as exploit to get unlimited treasury. The most usual culprit is Moscow, by the year 2029. IMHO this bug should not have escaped testing. Apart of this bug, I found game stable, if resouce hungry.
To sum it up, even as I highlighted some negatives, my overall feeling of the game is positive and - as I bought it for 22 euros and now just added to my steam collection for 2.5 - one of the best values for a buck despite its flaws. I would not hesitate to recommend buy for any global 4X games fan, who can handle limits of the game. And I hope Batlegoat will continue to improve the engine as Paradox did with Europe Univrsalis or HoI.