It is a tricky question..
See given in the language L1 we have |x| = |w| = |y|
But in language L we do not have the w component .Only x and y components are there..
And we require |x| = |y| as L is defined in terms of L1 in the question..
So as |x| = |y| , thus it is guaranteed that any string belonging to language L is an even length string as then only |x| = |y| is satisfied..
We know we can make a DFA of 2 states for the even length strings with say Q0 state corresponding to even length strings and Q1 for odd length strings..
Alternatively , we can have regular expression as : [(a + b)2]* = (aa + ab + ba + bb)*..
Thus the language L is regular..
Hence option A) is correct ..