Alright. Factors not really being considered or left out.
Firstly, offense set only gives 15% of offense before you count in offense from mods. Crit damage takes into account all offense from mods. However 30% crit damage increase is not a 30% increase in the damage you inflict when you crit. With no triangle, it is only a 20% increase in damage. with a 6e triangle it is actually a 222/192 = 1.15.. 15% increase in your crit damage.
Since we are talking about maximizing damage with offense or crit damage here, we can assume 3 offense secondaries and one crit damage with the offense set. Offense primaries will give 8.5% offense meaning 3 of them will be a 25.5% increase in offense. This means the amount of offense we get from the offense set will be 1405/1255 = 1.11... an 11% increase in offense This is before offense secondaries are taken into account, which will only decrease the value of the offense set. The primaries not being counted would increase the value. As it is, the break even crit chance point for crit damage vs. offense before secondaries are taken into account is 11/15 = 73% crit chance.
So, counting a crit damage triangle, the breakpoint where crit damage sets will do better than offense is now around 75% crit chance unless you have amazingly high offensive boosts in mod secondaries/primaries.
This value will decrease as more offense secondaries are added, but increase as crit damage bonuses are added in comp.
So, all relevant math included new guideline is around 75% final crit chance (counting mods and leader/composition abilities) is when you start using crit damage over offense.
