What does the output of the correlation look like in the typical experiments testing Bell inequalities when using light and polarizers? I am curious to know what a plot of polarizer angle vs counts produces for experimental results of Bell tests vs the classical prediction.
 A: You can't go too wrong by looking at the OG Bell's inequality test:

Experimental Test of Bell's Inequalities Using Time-Varying Analyzers. Alain Aspect, Jean Dalibard, and Gérard Roger.  Phys. Rev. Lett. 49, 1804 (1982).

Figure 4 from that paper looks like this:

Note that there is not one particular "classical prediction" to compare this to (which is why the plot only shows the quantum prediction.) Rather, Bell's inequality says that the correlations between various measurements must satisfy certain inequalities if local hidden variables exist.  Different hidden-variables models would presumably produce different predictions, each obeying Bell's inequality.  Aspect, Dalibard, & Roger found that their results were incompatible
with this inequality (see the top of p. 1807 in the linked paper.)
A: The law of Malus looks remarkably like Michael Seifert's graphic from Aspect et al.  Bell's assumptions excluded wave theory, therefore it has never been considered as a classical contender. All the recent experimental tests approximate the law of Malus and the quantum prediction. the the law of Malus and the quantum prediction are identical.
