The possible hairsplitting I see:

Half-life is defined properly only for one decay, not for a chain. But here, one decay is much slower than all others. Also, ^{206}Pb is the only final product. So, besides naming details, I'd agree with the proposed figures.

Decay is random. With moles, the figures are well defined. With only 1000 atoms, they are not, due to the very nature of radioactivity. Hesitating between 62.5 and 63 results from that. The standard deviation on 62.5 should be sqrt(62.5) or nearly 8, and 3 s.d. are not so uncommon, so figures between 38 and 86 can be reasonably expected. A more precise prediction like "63" would probably fail.

Here I computed as if distributions were Gaussian, but even this is wrong with such small figures.