A problem on redox reactions invites the reader to balance the actual reaction from which the redox simplification arose. After some hunting I found the balanced equation and attempted to re-derive the coefficients using:
$$\ce{a H_2SO_4 + b KMnO_4 + cH_2O_2 -> dK_2SO_4 + eMnSO_4 + fH_2O + gO}$$
We can quickly derive some valid relations by looking at each element:
$$ \begin{array}{cll}\\ \ce{H}&\quad a+c &=& f\\ \ce{O}&\quad 4a+4b+2c&=& 4d+4e+f+g\\ \ce{S}&\quad a&=& d+e\\ \ce{K}&\quad b&=& 2d\\ \ce{Mn}&\quad b&=& e \end{array} $$
This can be massaged a little but because there are 5 equations and 7 unknowns it seems that the system is underdetermined.
If we let $a=3,~b=2,~c=5,~d=1,~e=2,~f=8,~g=10$ the equations all work.
My question is:
Could we arrive at these values without guessing or actually doing an experiment to measure (say) the oxygen evolved?
FWIW, I notice an old paper - J. Am. Chem. Soc. 1889, 11 (6), 94-98 - of which I can only see the first page seems to take a similar approach and gets two possible equations, maybe for the reason above.