The results I'm getting from three different symbolic maths applications do not match my expectations. I have checked the result carefully and I still cannot explain the discrepancy. I half expect someone to find a flaw in my reasoning, because surely these three battle-tested applications cannot be wrong.
To make things simple I will be dropping the denominator in the rest of this post.
In all the applications that I have tried
$$ \int_{-1}^1 \int_{-1}^1 \int_{-1}^1 x^a y^b z^c \; dx \; dy \; dz $$
evaluates to zero, when $a = 2$, $b = 1$ and $c = 3$. Why this is so I cannot explain.
Here is my reasoning. If
$$ \int \int \int x^a y^b z^c \; dx \; dy \; dz = x^{(a + 1)}y^{(b + 1)}z^{(c + 1)} $$
Therefore the definite integral over the limits $[-1, 1]$ ought to be
$$ 1 - (-1)^{(a + 1)}(-1)^{(b + 1)}(-1)^{(c + 1)} $$
Therefore when $a = 2$, $b = 1$ and $c = 3$ the definite integral ought to be 2, not 0.
Furthermore, if it is true that
$$ 1 - (-1)^{(a + 1)}(-1)^{(b + 1)}(-1)^{(c + 1)} = (-1)^{(a + b + c)} + 1 $$
Then the definite integral ought to be 2 if $(a + b + c)$ is even.
The integrations should be performed separately.
$$\int_{-1}^1 \int_{-1}^1 \int_{-1}^1x^ay^bz^c\;dx\;dy\;dz$$
$$=\tfrac1{(a+1)(b+1)(c+1)}(1-(-1)^{a+1})\cdot(1-(-1)^{b+1})\cdot(1-(-1)^{c+1})$$
That's not the same thing as $$\tfrac1{(a+1)(b+1)(c+1)}(1-(-1)^{a+1 + b+1 + c+1})$$ which is what you appear to be computing. It looks like you are integrating each factor and evaluating the product of the antiderivatives. Nope.
In other words, $$\left[f(x)g(x)\cdots\right]_{x_0}^{x_1}\neq\left[f(x)\right]_{x_0}^{x_1}\cdot \left[g(x)\right]_{x_0}^{x_1}\cdots$$ in general.