John Gray was the last of the permanent teaching staff to present at our first year Government department seminar. Rather than outline a political project of his own (which I was told afterwards he probably wouldn't agree with), he outlined the two main approaches that liberal theory can be studied: either in a historical context (whereby each contribution to the liberal cannon must be understood within the specific context of the time) or in an ahitsorical, analytical way (i.e. looking at the theory on its own merits). There are challenges presented through both and reflect a trade-off between the two.
Although I'm no political theorist, I'd say that these tensions exist in political science as well. You can have the most perfect theory imaginable, but it needs to be tested against historical cases. Likewise, history on its own is descriptive, but doesn't offer any theory or pattern; instead it's just a story.
Not that this observation, either by me or Gray, is profound. But it's worth reminding ourselves about this from time to time.