Actually I have chosen not to even work on this because I doubt that it is true.
I say that because I have never see any problem even close to this one.
Thank goodness I'm not a "true PhD mathematician" to be dismissive of such a problem. Rather, I knew to have an option
to plow ahead and test a = b = c, for starters. That comes up with the inequality \(\displaystyle \ 3/2 \ \ge \ 3/2. \ \ \) That's promising.
Then, I may look into letting b = a, and c = ka, where 0 < k < 1 for one case and k > 1 for the other case
(so that I have two variables equal to each other, and the third variable is distinct from them).
From that I got it down to \(\displaystyle \ (2 + k)^3 \ \ versus \ \ 27k. \ \ \) Someone may plot \(\displaystyle \ y = (2 + x)^3 \ \ and \ \ y = 27x\)
on the same set of axes to help see the cases mentioned above.
This continues to look promising.
But then there is the \(\displaystyle \ a \ne b \ne c \ \) case still to work with . . .
- - - - - - - - - - - - - - - - - - - - - - - - - - - - -- - - - - - - - - - - - - - - - - - - - - - - - - - - -
I am supposing that this inequality, if true, was not proved by cases as I am hinting at.