I don't believe I have ever established some % of the populace that must either actively or passively support the insurgency, only that it is a significant segment of the populace. Many will want the status quo, many will want change, but change that is far different than what the group that has opted to add a violence LOO to their Ways and Means.

Even in a Democracy, a Bill Clinton can become President on 43% of the popular vote. Doesn't matter that 57% of the populace voted against him.

If the segment of insurgent populace is too small, too insignificant, like the Tim McVeigh crowd in Montana, it really doesn't rise to an insurgency as they are just too fringe of an element and the masses agree with neither their platform nor their tactics.

But when a significant portion of the populace agrees with the platform, even if they disagree with the tactics and don't actively join the insurgent organization you have an active populace based insurgency on your hands that must be addressed at the causal roots.

But to simply view an insurgecy as a military target to be engaged and defeated is very, very dangerous indeed; and typically merely provides some temporary sense of satisfaction for the counterinsurgent ("there, that will show them!"), while actually expanding the grievance and popularity of the cause, thereby making the insurgency stronger.

Firing rockets at Taliban camps in Pakistan makes Americans feel good. Motivating previously separate bands of the Taliban into a unified force against us and the government of Pakistan while increasing the percentage of the local populace's support for them is a very bad, (and very foreseeable) resultant consequence...

Sending the Worlds best Army to Boston, or the largest fleet in the world to New York may make the British feel good, but again, it was the wrong tool applied to the wrong problem to actually resolve the insurgency in a manner likely to be favorable to the counterinsurgent.