I don't see a clever way to invoke Ceva, Menelaus, or Stewart here. Nevertheless, there's a general principle at work. Consider $\triangle ABC$ with points $D$, $E$, $F$ on appropriate sides at shown ...

... where we define
$$p := \frac{|\overline{BD}|}{|\overline{BC}|} \qquad q := \frac{|\overline{CE}|}{|\overline{CA}|} \qquad r := \frac{|\overline{AF}|}{|\overline{AB}|} \tag{1}$$
(These are not Ceva-Menelaus ratios.) Then, for instance, since $\triangle AEF$ shares an angle with $\triangle ABC$, but the corresponding sides enclosing that angle are scaled by $(1-q)$ and $r$, we can write
$$|\triangle AEF| = (1-q) r\;|\triangle ABC| \tag{2}$$
Likewise,
$$|\triangle BFD| = (1-r)p\;|\triangle ABC| \qquad\qquad |\triangle CDE| = (1-p)q\;|\triangle ABC| \tag{3}$$
so that
$$\begin{align}
|\triangle DEF| &= |\triangle ABC| - |\triangle AEF| - |\triangle BFD| - |\triangle CDE| \\[4pt]
&=|\triangle ABC|\;\left(1-(1-q)r-(1-r)p-(1-p)q\right) \\[4pt]
&=|\triangle ABC|\;\left( 1 - p - q - r + p q + p r + q r \right) \\[4pt]
&=|\triangle ABC|\;\left(\; (1-p)(1-q)(1-r) + p q r \;\right) \tag{4}
\end{align}$$
Observe that $(4)$ is obviously unchanged under the substitutions
$$p \leftrightarrow 1-p \qquad q \leftrightarrow 1-q \qquad r \leftrightarrow 1-r$$
This implies that,
If $D$, $E$, $F$, $D^\prime$, $E^\prime$, $F^\prime$ are such that
$$\overline{BD} \cong \overline{D^\prime C} \qquad \overline{CE} \cong \overline{E^\prime A} \qquad \overline{AF} \cong \overline{F^\prime B} \tag{$\star$}$$
then
$$|\triangle DEF| = |\triangle D^\prime E^\prime F^\prime| \tag{$\star\star$}$$

(Note: To avoid marking overlapping segments, the diagram depicts
$$\overline{BD^\prime} \cong \overline{DC} \qquad \overline{CE^\prime} \cong \overline{EA} \qquad \overline{AF} \cong \overline{F^\prime B}$$
but clearly these conditions are equivalent to $(\star)$.)
For the problem at hand, one needs only show that the points of contact of $\triangle ABC$'s edges with its incircles and excircles make a collection of points $D$, $E$, $F$, $D^\prime$, $E^\prime$, $F^\prime$ satisfying $(\star)$. Well, this certainly looks true:

Proof is not too difficult. See, for instance, the first part of this answer. $\square$