Red on Green is Readable
Red on green is readable, including by people with red-green blindness, it just depends on the shade of red or green (This is #DC0000 on #82FEA9).
And we can predict it will be readable because the brightness difference, which is -141, is above the breakpoint for brightness difference, irrespective of hue, so even if you have red-green blindness it will be readable. (Try it on http://colorfilter.wickline.org).
We could convert the same brightness levels to monochrome, so as to simulate how someone with no colour perception at all would see it. This is #424242 on #CFCFCF, which also has a brightness difference of -141. (It has a W3C colour difference of 423, so should be out of readability range according to the W3C guidelines, in contrast to the red-on-green sample above which has a colour difference of 513 so should be within readability range.)
Looks fine doesn’t it?
Notice that the ‘Colour Difference’ in monochrome is below the supposed readability level, and in colour it is above the level, which yet again indicates what nonsense the ‘Colour Difference’ measurement is. Who came up with that stupid idea anyway?
Brightness difference is the key to readability, and it is so significant, that you can ignore the issue of designing for people with colourblindness. If the brightness level difference indicates that the text will be readable, it will be readable by people with colour recognition deficiencies, because there’s adequate difference in brightness, irrespective of the hue.
Where there is inadequate difference in brightness for the text to be considered readable, that doesn’t mean that the text is unreadable, but it may be to varying levels hard to read. There’s more on this on my Colour on Colour Readability page.