They probably don't correspond to Wratten numbers at all. The experimentally-derived filter factor for the red filter, for instance, is 2, whereas the filter factor for a #25 is 5, so there's an extra stop and a third of light coming from somewhere, and that somewhere is probably in the shorter wavelengths. The other filters are quite similar. That suggests that the Q of the filters (the slope of the cutoff curve) is rather lower than those of the old standbys.
Your best bet is just to experiment. If they were Wratten filters with gummint-approved bandblocks and bandpasses, you'd need to use a photospectrometer to find the -3db point and the Q of the filter to figure out exactly which of several similar filters you might be dealing with (or you can get a good idea eyeballing them side-by-side with known filters and a flat-spectrum light source). But whether you use lab test equipment or known-good filters, it's probably more expense than it's worth for a set of $4 filters. And unless you're doing colour separations, the Wratten number is really just a shorthand way of saying "I know what this filter does", and that's a body of knowledge that you can easily derive from a roll or two of film. And who knows? You just may find that these inefficient filters are actually better for many types of photography than the "real deal". (The red, if it's leaking enough green and into the blues, would probably still be "good enough" for dramatic skies while being almost ideal for women's portraiture in the Hollywood glamour vein.)