I think the MSC. Fatigue is better since their products offer more features than other relevant software, like MSC. Marc and ABAQUS (which I'm currently working with them). However, I think it depends the most on your demands.
You can also learn some general features of MSC. Fatigue from the following link in YouTube:
In 2013 Dassault Systemes acquired the British company Safe Technology and consequently incorporated the fatigue software fe-safe to its portfolio. Now fe-safe is being looked after by SIMULIA, meaning that ABAQUS and fe-safe are closer than ever. With the tolkens system you may have enough credits to run fe-safe if you're already an ABAQUS user. Just a suggestion.
I think MSC. Fatigue is better for the following reasons: 1- MSC. Fatigue is more user-friendly. 2- There are more available methods to calculate fatigue life.
3- Time History Rainflow Cycle Counting Both the traditional Stress-Life (S-N or total life) and Strain-Life (E-N, local strain or Crack Initiation) methods are available.
4- With minimal knowledge of fatigue analysis, users can perform such evaluations directly in their familiar FE modeling environment.
FEMFAT software has a great advantage over MSC-FATIGUE which enables you to have stress-based spot weld analysis while the spot weld module of MSC-FATIGUE requires some special model, loads, and etc. Moreover, there is a very good feature in FEMFAT which is spot weld remesh where you can convert mesh independent model to mesh dependant model in a few clicks. Besides, the fatigue criteria in FEMFAT is significantly lack and there are just some stress combination while there is wide range of criteria in MSC-FATIGUE from famous uniaxial one such as Goodman and Gerber to multiaxial ones such Fatemi-Socie. Overall, FEMFAT is more suitable for industrial applications where consistency is more important while MSC-FATIGUE could be better academic option due to its Fatigue criteria.