First, numerology is hardly the path to understanding science. It's beloved by crackpots who spend a great deal of time finding simple, or in some cases complex, wholly unmotivated relations between measured quantities. Indeed, in many cases even when better measurements show that these relationships are spurious, the proponents cling to their ideas.
Second, your calculation gives 87.381 +/- 0.004 GeV. The measurement is 80.398 +/- 0.025 GeV. That's 280 standard deviations away, hardly "almost". With such a large number of possible expressions, surely you could have done better than getting within 9% of the measured value. Why 4 pi^2? Why not (283/45)^2?
Third, electroweak radiative corrections, mostly due to the top quark, pull the mass of the W up. Without them, the W would weigh 77.5 GeV. That makes your numerology even worse.