The article subtitle says "Terraforming Red Planet going to be much harder than thought". This is erroneous, since the background loss rate (100 g/s) divided into the mass of the atmosphere gives a 7 billion year life. When combined with solar storms, we still get a "half life" of 500 million years. Half-life is the appropriate measure, because the loss rate is proportional to how much atmosphere there is to lose. If we can terraform the atmosphere in the first place, we can certainly compensate for leakage of 1% every ten million years.
The idea of terraforming the whole planet, though, is not reasonable unless you have many millions of people living there. It is more sensible to terraform only the area under your habitat domes, where the people are. Since outside pressure of the Martian atmosphere is much lower than Earth-normal pressure, the dome is a "pressure vessel" in the engineering sense. It turns out the minimum load is when the weight of the dome is equal to the pressure inside. On Mars that is 25 tons per square meter. It doesn't matter what you use to weight the dome, glass or rock. But when you put that much matter overhead, it is effective radiation shielding, even better in fact than the 10 tons per square meter of our Earth's atmosphere.
As the population grows, you keep building more domes, until you dome over the entire planet. At that point you have solved both the radiation and atmosphere leakage problems.