It was the time to step forwards, but...
SMAC heirs the features that made Civilization a classic, being the most important replayability: a random map makes up for that, and so does adding different factions with unique abilities; personally, mission-based titles aren't in my preferences. Diplomacy, although limited, has very well configured messages to give the feel of talking to a leader with his/her own personality; add to that agendas and aversions, and the Council, too. The gameplay in general is also a strong point: like in Civilization, found cities (bases), terraform, expand at will in a terrain with infinite variations and landscapes. The finest addition to politics is the Social Engineering, with great Role-Playing stile statistics.
Some say the technology tree is confusing due to strange descriptions: I think those add a unique sci-fi flavor: personally, I get more immersed in the futuristic atmosphere when reading "Our scientists discovered Photon-Wave Mechanics" or "Pre-Sentient Algorithms" than reading "Our scientists discovered the Photon Wall armor" or "can now build the Hunter-Seeker project".
The game is also very configurable, and multiple paths to victory enrich the gaming experience. Now, instead of having pre-made units, you can design your own, but shortly becomes evident which are the best combinations for each situation; the possibility is nice, though.
Graphics and sound aren't outstanding, but make their work, and that's enough for a strategy TBS game; it doesn't need more.
In general, SMAC is an empire-building game, and that's the feel it gives the player: of being building (or falling with) an empire through time.
However, SMAC also heirs bad and annoying features from Civilization. And that's the matter: Civilization is a classic, with no doubt. But it was released in 1991, eight years before SMAC; and that's too much time. Many fans, I included, were disappointed by this game because hardly can be referred to as "evolutive".
First, diplomacy is still limited: instead of constructing your own proposals and counterproposals, the game forces you to take pre-defined options (what can I do when I want to demand or give 457 credits and the computer only allows me the fixed number of 300 or 150?). And diplomacy problems brings treaties and alliances problems: there's no way for a treaty to stand for long: if you're weak, the strong empires crush you; if you're strong, all other factions declare war on you even if they have no chance (apart of slave factions, of course).
Nothing to say about trade: the model is simple to death. "Tired of designing, guys?" is the question.
Developing of cities, also from Civilization, quickly becomes boring and repetitive: except on early stages of the game, it's a matter of "all buildings you can build". I think a model close to MOO1 or perhaps Ascendancy is much better. The "Talent/Drone" model, also, is too simplistic: does not include, for example, the effects of a sudden change of government (apart from varying Police indexes). What if a long-standing Democracy turns to a Fundamentalism in a turn? Will citizens remain so conformist? And so will they if the base is starving to death? And what if an obscure technocratic secret society is turning a base against its Eudaimonic government, or communists/ecologists incite uprising in a Free Market society? How can citizens stand quiet when unit after unit is destroyed in an endless war? Are soldiers mindless androids? Balancing Talents with Drones is enough to maintain peace?
But the worst of all is combat. Perhaps seems combat is not very important in an empire-building game, but is impossible to play a game without messing in several unprovoked wars (even with long-term allies; perhaps they don't know allied victory is possible). The "collateral damage" feature is annoying to say the least: I don't understand why 7 units sit while the 8th. is fighting against 1 attacker, who inflicts damage to all when winning; also, "automatic retreating" of tanks is a bad fix. Why not using a system like MoM? But even worse are the Attack/Defense stats: a battle consists of the attacker firing at the defender while the latter stays hoping its armor deflects all shots. A 12-1-1 Unit attacking a 16-4-1 Unit has all the chances of winning! Ridiculous! To make things even worse, armor values increase more slowly that weapons, making wars a matter of sending 13-1-1s. against 1-6-1s. in a base equipped with Perimeter Defense: outdoor battlefields belong to the one who attacks first. Air combat is still bad: although it's resolved comparing weapon forces, armor doesn't count. Why a land/sea unit cannot enter a square occupied by air unit?
Finally, although the AI is decent (decent, but not excellent), difficulty means more drones and ultra-cheating. The computer seems to know exactly where your units are, even without entering your datalinks or without having the Empath Guild. Do I really have the chances of success presented in the probe team operations window? No, I think. How can the computer discover Unified Field Theory and on the next turn send me a horde of marines equipped with weapon 12? He had only 150 credits to buy!
The Bottom Line
SMAC has almost all good features from Civ, and good new ones, but also has bad features from Civ (I say again: 8 years is enough time to get new and good ideas), and lacks some from Civ that were good.
Finally, even with the weak points, the game is worth a try, specially at reduced price. But i'ts far from being a true masterpiece.