I cannot do this by crossing my eyes (focusing on a point between you and the image), I have a hard time getting the cross to stay consistent and it never really "locks in" for me. Instead of crossing my eyes, I unfocus them, effectively look through the image. Once I get the repeating part to overlap cleanly, after a second or two, my pupils adjust their focus and the image fades from blurry to clear in a really satisfying way and kind of "locks in" in a way that takes little to no effort to maintain. With a bit of practice, I can even move my eyes around and look at different parts of the two overlayed images without distrupting the effect at all.

I don't know if it's just my brain working differently or if a there is some confusion in the discussion between crossing your eyes and focusing through an item.