For 1, it is incorrect.
For example, if $S=\mathbb N$, $X_0=S\setminus\{0\}, X_1=S\setminus\{1\}$.
Now, if accidently $f(S\setminus X_0)=\{S\setminus X_0, X_0\setminus\{1\}, \{1\}\}$ and $f(S\setminus X_1)=\{S\setminus X_1, X_0\setminus\{0\}, \{0\}\}$, then $f(S\setminus X_0)=f(S\setminus X_1)$.
To show that $|\mathcal P(S)| \leq |B(S)|$ we need to be more careful, a way to ensure it is to pick the "$a$"s more carefully, e.g.:
First let $(s_i\mid i\in |S|)$ be some ordering of $S$
If $|S\setminus X|>1$, do the same as you did, if $S\setminus X=\{x\}$, choose the $a$ to be such that $a>x$.
This is injective: It is clearly injective for elements $X$ such that $|f(X)|=2$, take $X_0,X_1$ whose image is of size $3$ and are both of cardinality $>1$, if $f(X_0)=f(X_1)$ then $\{X_0, (S\setminus X_0)\setminus\{a\},\{a\}\}=\{ X_1, (S\setminus X_1)\setminus\{a\},\{a\}\}$, from cardinality arguments $\{X_0, (S\setminus X_0)\setminus\{a\}\}=\{X_1, (S\setminus X_1)\setminus\{a\}\}$, but because $a\in S\setminus X_0,S\setminus X_1$ we get that $\{X_0, S\setminus X_0\}=\{X_1, S\setminus X_1\}$, so (by the way we arrange this kind of tuples) $X_0=X_1$.
If $X_0=\{x\},X_1=\{y\}$ for $x\ne y$, and $\{y\}∈f(X_0)$, then $y>x$ which means that $\{x\}∉f(X_1)$, so $f(X_0)≠f(X_1)$.
As for 2: I claim that $|B(X)|=|\mathcal P(X)|$:
Again, remember that $(s_i\mid i\in |S|)$ is a well ordering of $S$, now define an injection from $B(X)$ to $S^S$:
If $π∈B(X)$, for $x\in S$ we denote $[x]_{π}$ to be the unique $ζ∈π$ such that $x\in\zeta$ (the projection of $x$ onto $π$).
Define $f_π:S\to S$ as follow: $f_π(x)=\min ([x]_{π})$.
Let $h: B(X)→S^S$ defined as $h(π)=f_π$.
$h$ is injective: if $π_0\ne π_1$, let $ζ∈π_0$ such that $ζ∉π_1$ define $η=[\min(ζ)]_{π_1}$, if $\min(η)≠\min(ζ)$, then $h(π_0)(\min(ζ))≠h(π_1)(\min(ζ))$ hence $h(π_0)≠h(π_1)$, if it is equal, take an element from $ηΔζ$, this element will witness that $h(π_0)≠h(π_1)$.
So we get $|\mathcal P(S)| \leq |B(S)|≤|S^S|=|\mathcal P(S)|$, so they are all equal.
\text{ }to write text in the middle of a math expression. Look at my edit of your question. – jjagmath Apr 23 '22 at 12:34