While it's fun to think about battles of yore with waves of tanks sweeping in grand formations across the steppe, the reality is modern wars are fought somewhere on the same scale as repression, occupation, political movement, and large policing action.
This sort of conflict has always existed, and was perhaps the only type of major conflict until laws of war were formalised. Even during World War 2, partisan activity tied down significant resources in Europe, and many political revolutions were born of this sort of conflict at the start of the 20th century. This is why I prefer the term Asymmetric Warfare to Fourth Generation Warfare.
Those two terms do actually have different meanings and connotations, and Fourth Generation Warfare is worth talking about in the context of western powers. Technically, all warfare is asymmetric to some degree, but in the context of asymmetric warfare at least one faction in the conflict has very little de jure authority or de facto military power compared to its opponent. The other faction may be the current government or local military (its legitimacy is of no concern for the time being), or an occupying force from another country.
If we use that definition (which is as broad and neutral as I can make it), one can see how it relates to modern and contemporary political movements. The rules of the conflict are different and thus the strategies present are different. 'Fighting' protesters in a democracy is very different to fighting tribal militias in the hills. But, ultimately, they are both about fighting over people. The staggering human cost associated with figuring this out in structured militaries is a testament to the complexity of those conflicts.
A compliant local population (for the purposes of this, the local population is the population of the region where the conflict is occuring) is relatively cheap for the occupying force, and really the ultimate goal of such a conflict. Without support from at least a portion of the population, the small insurgent force will run out of both those willing to support it, and those willing to equip it. As modern technology improves, the portion of the population goes down, but it's still quite high. Let's say the homosexual emancipation movement as a "break-point", around 40-60% support for gay rights giving a fairly solid bloodless advance on their objectives. In the context where people are getting shot regularly for open support, I'd imagine there being more casualties on all sides but ultimately less support is needed for regular supply (the sunk-cost fallacy in action).
For the occupying force, unless they are willing to commit to genocide, they must check their power. During World War 2, there was a theme among many of the democracies that the war could be won through demoralising their enemy's civilian population, while propaganda would strengthen their home front. However, it was discovered that the brutal treatment of the local population hardened them and gave them more reason to continue fighting (who knew?).
In a more contemporary context, the occupying force must be seen as a stabiliser for the region, providing security through necessity. This itself is still not a learned lesson; the USA contracted a lot of US companies (duh) to rebuild and upgrade the infrastructure of Iraq, rather than employing and paying locals to do it. The behaviour of the contractors, while shitty, is not the fuck up. Once people have livelihoods and those livelihoods are at stake, they are much less likely to support insurgents. Obviously there is a security risk there, but it is less likely to breed bad blood.
That is one instance of a major failure that could have been easily avoided. Different regions will have different rules (regardless of how dumb you might find them to be). But learning from others' mistakes is probably the cheapest way to learn (though one must take care in understanding the context in which they made those decisions), and there are many common themes.
For the occupiers, whether or not they are actively exploiting the local population, they must appear as though they are not. There are reasons for occupation beyond the more cynical ones (as many of my friends believe), but the reasons are mostly irrelevant. Just to throw out an example, there is ample reason to move troops into an unstable neighbouring region to install a government (as legit as you want) aligned closely to you. A lot of the rules for the occupiers can be boiled down to "don't be a dick".
For the insurgents, it is important to make the occupiers appear cruel, callous, or both. But it is also important that their main 'bases' or hubs remain relatively hidden or innocuous and separated, such that they cannot be targeted with any degree of alacrity. Obviously, the second point is less important in conflicts that don't use explosives. There is very little thought of individual conquest; while ISIS talks big, it is unlikely that they will directly cause the collapse of NATO.
I'll probably post more about specific situations while gathering thoughts for more general rules. Fighting in such a conflict is likely to cause significant psychological trauma for everyone involved, and I don't wish to marginalise that. But in a scrap only one side comes out on top, and in the moment that's all that matters.