You're missing the main thrust of their effort. It's not to individually tailor ads, it's to study consumer behavior to the point where they discover what actions they can take to cause the effect they want.
The idea is similar to the reason gas stations have loyalty cards... they're not so loyal customers get discounts, it's because they can sell that data to companies wanting to analyze spending habits.
If Google can use its expertise to cross-correlate data from financial services like it does other sources, that fills in a link in the chain, from money going into the consumer's pocket to money coming out in a purchase.
Gas stations already know that e.g. (making this example up) someone who buys gas at a certain time of the morning has a X% chance to also buy a breakfast donut if they buy a coffee. They get this from statistical analysis of the data from loyalty cards.
So all the gas station has to do if they want to make some amount of money from the people who buy donuts is to trigger the behavior - provide free coffee or coffee at a discount at a certain time of the morning, and at the same time add a few cents to the price of donuts. In exchange for a loss on the coffee, they get a very high likelihood of making extra money by manipulating consumer behavior, and likely the consumers don't even know it. That percent advantage is worth dollars.
Google wants to correlate consumer data to the point where they know all the things that trigger when people spend. Other corporations will pay gigantic amounts of money for that service, because it's a bankable return for them. There's no guessing whether (going back to my example) free coffee in the morning will cause better sales of donuts, because they know as opposed to guess.
Google filling in all the blanks in its chain of data sources means that they'll be able to correlate things like "During a three paycheck month, men age 30-45 tend to put that extra disposable income in their checking account for use on a mix of necessities and luxury items.". If a company selling luxury goods wants to know when they should saturate men age 30-45 in a given market with ads to drive sales temporarily up 65%, Google can sell them that information.
Just knowing when specific people will have the ability and interest in buying something they're selling lets them avoid spending money on advertising, shipping, or even producing goods that won't easily sell. That's a huge advantage in the market over another company that has to guess when and what kind of advertising will be effective.
The consumer targeted by this will not know... they'll just be more likely to buy a particular product when they happen to have the money to do so because the sellers will be subtly increasing the coverage of ads or targeting those ads to the demographic involved during that time period, and those ads can manipulate as well using sex, rhetoric, or any of the other techniques advertising knows about.
Being more certain (even if not 100%) what will happen in the future translates into money and power.