This is not true. Morphs are mutations. Mutations can be defective but they can also be extremely beneficial... think X-Men. In all seriousness though, mutations are what drive evolution, can be beneficial or defective, and viewing them as a race to defectiveness is a poor way of viewing incredible morphs.
It is extremely unlikely but it is not impossible that adding a second mutation can cancel out the harmful effects of the first. Two ways that come to mind:
1) The Desert gene mutation creates a Stop codon which ends translation and causes infertility. Adding a second gene creates a mutation that changes the stop codon to a different, harmless codon that does not affect fertility.
2) Some amino acids have several different codons that code for them. The desert gene mutates a single nucleotide in females that creates a codon that causes infertility. The second mutation changes another nucleotide in the same codon to either create an original pre-desert amino acid or a new one that doesn't affect fertility...
Both of these scenarios are extremely unlikely but it is possible for another mutation to fix a mutation. I don't think this is the case for the desert morph and hadn't planned on jumping in but I couldn't let the morph = defectives comment go.