We start with the following: $$ (x+z)^2 - x^2 = \sum_{n \geq 1} \frac{z^n}{n!} \frac{d^n}{dx^n}[x^2] $$
$$ (x+z)^2 - x^2 = z(2x+z) $$
$$ z^k(2x+z)^k = \sum_{n \geq k} Y^{\Delta}(n,k,x)z^n $$
Where $$ Y^{\Delta}(x) = \frac{k!}{n!}B_{n,k}^{x^2}(x) $$ Therefore: $$ (2x)^k [z^{n-k}]\left(1+\frac{z}{2x}\right)^k $$
$$ = (2x)^k [z^{n-k}] \sum_{j=0}^k {k \choose j} \left(\frac{z}{2x}\right)^k $$
$$ = (2x)^{2k-n} {n-1 \choose n-k} $$ Is this correct? I feel like this is very wrong.
Only the last calculation is somewhat wrong (see Hints below). Everything else is ok.
(Hint: $Y^{\Delta}$ has three arguments)
(Hint: The exponent of $\frac{z}{2x}$ in (2) is $j$, not $k$)
We conclude from (3)
We can generalise the problem from $x^2$ to $x^N$ with $N>0$ a positive integer. A nice by-product is, when specializing the result to $N=2$ we will derive a binomial identity.
Again, we know from this paper by V. Kruchinin, that \begin{align*} Y^{\Delta}(n,k,x) = \frac{k!}{n!}B_{n,k}^{x^N}(x) \end{align*}
It follows \begin{align*} \frac{k!}{n!}B_{n,k}^{x^N}(x)&=[z^n]\sum_{n \geq k} Y^{\Delta}(n,k,x)z^n\\ &=[z^n]x^{Nk}\sum_{j=0}^k\binom{k}{j}\left(1+\frac{z}{x}\right)^{Nj}(-1)^{k-j}\\ &=x^{Nk}\sum_{j=0}^k\binom{k}{j}[z^n]\left(1+\frac{z}{x}\right)^{Nj}(-1)^{k-j}\\ &=x^{Nk}\sum_{j=0}^k\binom{k}{j}\binom{Nj}{n}x^{-n}(-1)^{k-j}\\ &=x^{Nk-n}\sum_{j=0}^k\binom{k}{j}\binom{Nj}{n}(-1)^{k-j} \end{align*}
Comparing this result with (4) above we observe