I'm familiar with the formula - I had read Greenblatt's book on it some years ago. The formula recorded some impressive returns until the 2000s. But since then, as with many other quantitative value investing strategies, it has failed to perform.
There are two possible explanations for the recent underperformance. One explanation is that such strategies tend to go through rough patches which sometimes stretch a decade or more, and that we're just going through one of these patches. If so, the strategy should perform again soon. The other explanation is that the proliferation of computers has "arbitraged" away any advantage of such strategies. If this is the case, the magic formula would never work again.
It's very difficult to know which of these explanations is correct. Personally, I think they're both partially correct. In other words, I think the magic formula will work to some extent going forward, but perhaps not to the same extent as it did before the 2000s.
Finally, if you already invest in QVAL, I'd think carefully about how much to allocate towards the magic formula. QVAL and the magic formula have some similarities, so you may find yourself investing in the same stocks twice.