Probably yes, but I'd want something else to correlate that with before deciding for certain. If the numbers are 32-bit then the range could mean 0x00000000 to 0x0000ffff.

I think what is meant by "weight" is the relevant power of 10. So 1234 means 1*1000+2*100+3*10+4, and that's what it means by multiply and add, and the resulting value of 1234 in the relevant register will contain the value which you can then represent in binary with the method you've outlined.

I'm not clear what is meant by a division method, maybe something like:
while binvalue:
store digit (binvalue%10)+0x30
binvalue/=10
end while