You're right: this is a different interpretation of $\Delta$ (or $d$) than usual, because there's no natural sense in which $\mathbf{A}$ is changing. Similarly, people also get confused by the moment of inertia,
$$I = \int r^2 \, dm$$
because there's nothing about the mass $m$ that's changing. Now, you can always cook up a picture where everything is changing (as I do for $I$ here) but it might be clunky. For instance, to interpret
$$\mathbf{F} = \int P \, d\mathbf{A}$$
I could say that I'm totaling up the force by gradually looking at more and more of the total area, so the 'total area considered' really does go from 'none' to 'all' in steps of $d\mathbf{A}$. Similarly I could imagine $P = dF/dA$ as telling me how the force on some surface would change if the surface were to expand.
Neither of these interpretations are very compelling, but we use the $d$ notation anyway because it's useful; it suggests things we can do. For instance, I can also calculate the force by splitting the surface into tiny $dz$ pieces. In our notation it looks very natural; it's just splitting fractions,
$$\mathbf{F} = \int P \frac{d\mathbf{A}}{dz} \, dz.$$
Now $z$ is a perfectly good coordinate we can regard as changing; if we put an upper bound of $z_0$ then $\mathbf{F}(z_0)$ just gives the force on everything below $z = z_0$, and so on. It's by virtue that we can convert the $d\mathbf{A}$ to an actual differential (in this case, by parametrizing the area by $z$) that we use the differential notation for it too.