Some of the examples on tallying Overwatch score are not compliant with the rules as given on pg. 176. I'm pretty sure this is due to the rules evolving thru playtest/development and the example text fell through the editing cracks.
Take the rules here as gospel, and where the examples don't match, go with the rules.
So I'm really not sure how it's supposed to work. If it's just opposing hits, though, then what's the consequence of labelling some actions without opposing rolls as illegal? Is it just for flavour, then?
So you gain OS from:
Any matrix action where you benefitted from a hacking cyberprogram (+1 OS per action)
Maintaining hacked access (+1 OS/round for User Access; +3 OS/round for Admin access)
Performing illegal matrix actions (+1 OS per hit on the opposing roll)
So, for example:
If Mungo throws a Data Spike at a goon's gun, he gains however much OS as there were hits in the resistance test because it's an illegal action. 0 hits = 0 OS. 10 hits = 10 OS.
If he used the Exploit hacking cyberprogram to lower the Matrix Defense Rating for the target of the Data Spike, then that's +1 OS.
If he used multiple hacking cyberprograms, it's still only +1 OS, as the language says it's per "action modified" and not "per hacking program employed"
If he had previously hacked the goon's PAN and had gained access, he gains +1/+3 OS as well for the round, even though that didn't come directly from the Data Spike.