The article is interesting, but both methods presented are quite strange to me. The denominators are equal, the number being subtracted is smaller than one and the number being subtracted from is larger than one, so it's just instant subtraction, no method needed. If the number being subtracted were larger than one, I might have two steps (if the denominators were different and not multiples of two, three because I will have to equalize them) - subtract the fraction, then subtract the integer from the result.