In Piketty's model, r is return on capital. So that means it's the return on things like equities (and btw, interest is compounding on investments, don't know why you represented otherwise), real estate (Piketty is not overly find of the role of rentiers/landlords in the economy), machinery and so on. So the money accrued by the owners of wealth and machinery that is, essentially passive income, or income generated by capital, is r.
g is the growth of the economy. We calculate it as GDP, what he's interested in is basically people's combined incomes and economic activity that goes with it.
The premise that r > g is saying that people who have wealth, their return on that wealth will always be higher than the growth of the economy. If the economy grows 1%, the return on capital (remember it's all capital in the economic sense) will be more than 1%. Now this isn't true for all years in all places, but it has been shown to be true in the long run, which is how wealth is accrued quicker by those who already have wealth, and how inequality increases. Keep in mind Piketty's book is a massive tome.
And btw, comparing a rate that's calculated as %/year to another calculated as %/year is mathematically correct.
Your example is nonsense btw, it almost seems like you've confused r with g, because interest certainly compounds with capital, whereas with income, while you'll get increases over time, it generally isn't of the same magnitude and relates to labour, of which you have a finite amount.