@FeiLiu I run the process in windows under Cygwin and it works correctly, try using csh shell instead of bash (I couldn’t use it under bash but no problems with csh). The only problem that appears is that mt_prep writes some files with the required paths to import data on matlab and due to the way Cygwin works the paths didn’t match with the ones on windows. However, the solution is as simple as modify this files (I added some extra lines to mt_prep to do it automatically when the process is finished, I can upload it if someone wants it).
@bayzidul from my relativlely short experience using SNAP, Tops Split step is necessary because SNAP can only corregister images in the same subswath (if you want to make an interferogram using the whole IW swath width you need to corregister the three subswaths separately and use tops merge operator afterwards). Deburst is also necessary to merge the different burts, otherwise your images will have a nice black fringes a part from the duplicated pixels that will appear on the overlapping part of each burst.
For lat and lon files simply rename rslc files as FeiLiu indicated and it is necessary to do it even if you applied TC. Finally, you are right about Stamps export and corregistered stacks of at least 4 interferograms are required. However, ifgs have to be created individually because topo removal operator doesn’t work properly with stacks (see annamaria post where she explains it) and create a stack of ifgs before exporting to snap.
Personally, I’ve followed the workflow indicated by katherine at the begining but using the approach that annamaria explained to create ifgs and the modified scripts that FeiLiu uploaded. Doing so I achieved quite good results.