It's part of the compounding evidence towards the stress corrosion cracking cause. Whilst the lug area is slightly bigger on the nut end due to the extra material, it's not a massive % difference. There have been quite a few cars - mine included - that have survived a period of time with the bolt head end failed and only the nut end secure. That implies that the lug strength alone is more than sufficient (how would a component now bearing twice the load it was designed to, plus a chunk of bending due to the other side failing, not fail immediately?) By hand calc, both the lug sizes and therefore strength appears more than sufficient. The main difference between the two sides is one has a permanent hoop stress imparted on it, the other doesn't. My lug had very small corrosion indications on it which I appear to have weakened the bracket sufficiently for that side to fail well below it's designed load. Bit of background below for those interested. Main point is that the permanent stress than needs to be applied to a material is often extremely small (sometimes <1/4) compared to it's yield stress. Geekily, it is pretty interesting! http://www.totalmateria.com/Article17.htm