The statement at the end of the OP, that "*If the function $f(x)$ is concave and positive the inequality of the beginning is reversed*" seems to have a counterexample:

For $f(x)=x^{99/100}$ the inequality 
$$\Big(f(1)^{f(1)}f(2)^{f(2)}\cdots f(n)^{f(n)}\Big)^{\frac{1}{f(1)+f(2)+\cdots + f(n)}}+\Big(f(1)f(2)\cdots f(n)\Big)^{\frac{1}{n}}\geq f(1)+f(n)$$
is satisfied for $n=3$, but not for $n=4$ or larger. The violation is small, but within numerical accuracy:


    In[1]:= f[x_] := x^(1 - 1/100)

    In[2]:= N[Table[Product[f[j]^f[j],{j, 1, n}]^(1/Sum[f[j],{j, 1, n}])
              +Product[f[j],{j,1,n}]^(1/n)-f[1]-f[n],{n,3,5}],10]

    Out[2]= {0.001042561976, -0.001537906443, -0.005870494687} 

"Out[2]" is the left-hand-side minus the right-hand-side of the inequality, for n=3,4,5.