So recently, I asked this question asking if $\sum_{r=0}^n\sum_{s=0}^r\dfrac{n!a^{n-r}b^{r-s}c^s}{s!(n-r)!(r-s)!}$ was a legitimate way of expanding $(a+b+c)^n$
Multi-binomial theorem
When working in more than two dimensions, it is often useful to deal with products of binomial expansions. By the binomial theorem this is equal to$$(x_1+y_1)^{n_1}\dots(x_d+y_d)^{n_d}=\sum_{k_1=0}^{n_1}\dots\sum_{k_d=0}^{n_d}\binom{n_1}{k_1}x_1^{k_1}y_1^{n_1-k_1}\dots\binom{n_d}{k_d}x_d^{k_d}y_d^{n_d-k_d}$$
Now say I wanted to write out $(a+b+c)^n$ in this way. Would I use this theorem to write it out as$$\sum_{r=0}^n\sum_{s=0}^r\dfrac{n!a^{n-r}b^{r-s}c^s}{s!(n-r)!(r-s)!}$$
Which then I explained how the expansion would be achieved. Now here's my question: Why is this a legitimate way of writing out the expansion of the trinomial theorem? I understand that yes, it is, and I understand the multinomial theorem, but would somebody be able to give me a combinatorial explanation of the formula?
Why my previous question does not answer my question
The reason that this is is because I had only received confirmation that yes, what I had reached as an expansion of the trinomial theorem was in fact equivalent to my original expression, but had not clarified why.