When elections are won or lost based a few percentage points, then giving a 8.7% boost to a campaign can certainly sway the outcome.
Possible, but very unlikely. The effect has been studied and quantified by political scientists. I don't have a copy of a paper handy, but very roughly speaking, it's a matter of diminishing returns. The first doubling of money can sway the election, say 10%, then you double it again, and gain 5%, then double it again, and gain 2.5%, and by the time you've outspent your opponent by 16x you're barely moving the needle at all.
What they did find is that the candidates who are ultimately more popular with the voters turn out to have been more popular with the donors. So the politician who outspent/outraised his opponent by 4:1 and won the election did so because he was more popular.
Since the total possible amount of spending can never get above a certain threshold, it's actually more efficient to be a good candidate and spend the money you do have to get that word out, than to try to spend your way to victory while being an asshole.
There are exceptions, but the rule holds most of the time. Frankly if you're corrupt enough, you might as well just bribe the election officials or voting machine vendors rather than keep piling money into above-board campaign tactics that don't really yield returns.