It all begins with a ratings corporation. For instance, let's take Nielson, who are pretty much the go-to guys for ratings and almost certainly the people who'll determine if Toonami is successful (not to mention how successful). Now in order to determine who's watching what they select a bunch of people from across multiple demographics and "TV packages". They can't grab data from everyone because there's simply too much data, the infrastructure isn't in place to detect it in a lot of cases, and it breaks several laws. So it must be done voluntarily. The people selected typically run the gamut from a nuclear family of 4 to your grandma. From a middle aged single guy to a stressed out college student in a cramped dorm. White, jewish, black, asian, christian, atheist. If they can find a minority, they're represented in the sample as well because it prevents cultural bias. Frequently they'll toss out perks such as "free/cheaper cable" or a couple extra bucks a month to sweeten the deal. After they get about 5-10K of these people, they hand them out a small, roughly modem sized box. This box is wired into your cable and detects which channels you happen to be tuning in to. Once the survey is over, the data is collected and processed. They take the percentage of people who watched each show and then extrapolate how many people in the country are watching, as well as what particular subset of viewer they're in. For instance, a show might be a big hit with "the college crowd" or "a great family comedy". Maybe it's an "ethnic hit". This is then usually represented in ad campaigns and the like once the ratings are in.
Obviously it's not always super accurate. There's more than a few shows pulling in huge ratings that probably shouldn't be, whilst other wildly popular shows aren't being seen by anyone in the sample. I'd wager to guess there's more than a few shows that went off the air simply because the people chosen simply weren't the one's watching, whilst their similar social standing neighbors may have been. Popular petitions to bring back shows that seemingly had a fanbase big enough to support it also showcase this possibility. But it's really the best (and most legal) option.
The number of people watching then determines the ad cost. Basically, if there's a lot of people watching an event then you can charge more for ads. Meanwhile shows with low ratings in poor time slots have to offer bargains to get ad revenue. This is where shows (and networks) get most of their money. So it's always about trying to get the highest ratings across the boards. There's more than a few pieces of fiction (and even some non-fiction) about ratings tampering that are well worth checking out. It's a huge business and exploiting it can be highly lucrative. Which is also why a company like Neilson is regarded fairly well, since they work fairly hard to get as fair and untampered a sample as possible.
In short, you watching or not watching Toonami doesn't mean shit. If those five thousand people getting surveyed are the only people in the entire world watching TV, no one would be any the wiser. If they're the only people watching Toonami, it'll be selling adspace at the same cost as the Super Bowl. If none of them watch it, it'll probably be off the air before long. Because ratings are purely a statistical extrapolation through semi-random survey. The only way you'd be helping Toonami out would be by getting selected for a Nielson survey and going through with it. Then keeping your TV on that network whenever it's on. Since that's not the case, you watching or not watching doesn't earn them a dime.