nah its nothing to do with logs really
its like this
If you were plotting say the amount of volume (V) against time (T) for example and u were adding 1 cm3 each time,
V = 0 T = 0
V = 1 T = 1
V = 2 T = 2
etc..
at the end you find out that ur last volume (which took 6 seconds) isnt 5 cm3, its 6, because you put 1 cm3 too much in,
instead of plotting ur final point as V = 5 and T = 6, you simply use V = 6 and T = 6. You'll still get a graph which works and the mistake of adding too much in wont matter at all.
So similarily, you could have used V = 26 cm3 instead of 25 cm3 on your graph and you wouldn't have ruined the trend.
As far as reliability goes, its really up to how ur results were. There are two things with reliability usually. Firstly, do the points lie close to the line and secondly, were repeat readings done for each value done. I meant very reliable with regard to the first bit really. With the RDS, what i mean is that NaThio wasnt really involved in the main reaction (it was only being used to alter the concetration of the H2O2).