Strategy titles are of various kinds and scope. From the very first Civilisation masterpiece, to the more modern Imperium Galactica II, research is a common element managed by the game. Up to now, however, whatever the game style, the research has always been set in the same way. It is always a map of predefined objects, and you cannot deviate from that path.
Take Imperium Galactica II as an example. This strategy game is a wonderful accomplishment in its own right. A profound game, it can last a long time and no two games will play exactly the same. Nevertheless, its learning curve is not too steep, although a casual gamer has a very slim chance of finding it interesting. This is a game for hardcore strategists. Ones that do not mind spending three or four hours of an evening just to find out if a specific decision is worthwhile. Definitely not the game of choice for those who prefer a quick race around a track, easily started and easily forgotten.
Definitely not this game's type. IG2 makes you think down to the elements that you put in your ships. It is a game that allows you to personnalize your units, so you can have several of the same kind of hulls but they can contain vastly different weapons and characteristics. Unquestionably, IG2 is the game that gives you the most varied possibilities in creating your armada - which makes for even more differences in gameplay between sessions.
So why do we still have to suffer through set upgrades ? Why, if ships already accept differnt configurations for identical hull types, do we have to accept identical upgrades ?
I think, given the complexity of today's games, that it is time game developers create a flexible research tree. I wish for an IG3 that, in addition to existing functionalities, gives the user the possibility to design personal ship hulls, define possible content and research individual upgrades. How can this be possible ? I won't say it can be easy, but it should be less difficult than developing AI for soldiers in Half-Life and such.