So if the impurity is Iron(II) Sulphate that's not fully hydrated, say for example, it's completely dehydrated, then for the same amount of moles (obtained from the redox titration) your actual mass (ie factoring in the impurity) will be less than the mass you calculate, from doing the moles times by the Mr of hydrated Iron(II)Sulphate.
For the reducing agent point, the idea is that if the impurity is not able to reduce/react with dichromate then it won't affect the titration, as you're simply measuring the moles of dichromate that react until it's in excess, then using that to get a value for how many moles of Iron(II) sulphate must have been present. This is a similar idea to the fact that in acid/base titrations, you can put as much water in the conical flask as you like, as it doesn't effect the actual amount of acid/base in the flask, so the amount of reaction is the same.
Hence, if you've got an impurity that doesn't get involved in the reaction (ie doesn't reduce dichromate) then the moles of Iron(II) Sulphate you think you have, will be larger than the amount actually in the sample X.
Hope this helped!