1] Why not use result equivalence?
$\rightarrow$ Serial schedules need not be result equivalent.
Consider this simple example with 2 transaction T1 and T2 where x = 100 initially;
S1 : T1(x=x+100);T2(x=x/2) Result1 : x=100
S2 : T2(x=x/2);T1(x=x+100) Result2 : x=150
$\rightarrow$ Now if we go with result equivalence then we have to compare result of concurrent schedule with every possible serial schedule. This is hell of a computation.
Also, there is redundency in calculation. I mean, consider the fact that data item (x here) will have dynamic value. As value of x is not fixed, we will have to calculate result of every serial schedule everytime and compare it with concurrent schedule. Now, as we are computing every serial schedule why do it concurrently again??!
2] Can there be a schedule which has different order of conflicting operations but same result ?
$\rightarrow$ Yes, there is a possibility of this. (You think of the example ;-) )
It would be called as result equivalent but not conflict equivalent.