The threshold parameters for corrected bicarb and delta delta ratio are set to be slightly different. The delta ratio amplifies the degree of error with the formula as the measure bicarb decreases, since it is a ratio.

E.g. For a mixed acidosis + (metabolic) alkalosis, the delta ratio is usually said to be > 2, and the corrected bicarb > 30.

At a measured bicarb value of 18, the delta gap would need to be at least 12 to satisfy the delta ratio rule. For the corrected bicarbonate rule > 30, the delta gap would also need to be 12 or higher.

At the measured bicarb value of 12, the delta gap would have to be at least 24 to satisfy the delta ratio rule. On the other hand, for the corrected bicarb rule at > 30, the delta gap would only have to be 18 or higher.

Finally, at a measured bicarb value of 8, the delta gap would have to be at least 32 by the ratio method but only 22 by the corrected bicarbonate method.

As for why one and not the other, who knows? Our methods for these estimations are always slightly erroneous due to the assumptions we use, for e.g. we assume all buffering takes place in the ECF, and that HCO3 is the only buffering ion.