S1TBX co-registration issues

Dear S1TBX Developers,

I have encountered an apparent issue with the co-registration process to create a stack of S1A images over Indonesia, where the output stack is poorly co-registered. I have output the residuals text file and after 19 parses we have only 2 valid GCPs. What is interesting is that if exactly the same workflow is processed using the NEST toolbox (with the same inputs) the output stack is nicely co-registered. The NEST residuals text file reports 1678 valid GCPs after only 2 parses.

Here is a link to the coregister.xml graph:
ftp://l3-server.infoterra.co.uk/pub/projects/SAREDD/Sentinel-1A/work/coregister.xml

I have pasted below the command I am using to initiate the co-registration process:

/opt/S1TBX/gpt.sh
/data/Sentinel-1A/xml/coregister.xml -e
-PfileList=/data/Sentinel-1A/work/20150615_111742/S1A_IW_GRDH_1SDV_20141018T224822_20141018T224847_002890_00345D_AD3E_ml.dim,/data/Sentinel-1A/work/20150615_111742/S1A_IW_GRDH_1SDV_20141111T224821_20141111T224846_003240_003BE1_EEEF_ml.dim,/data/Sentinel-1A/work/20150615_111742/S1A_IW_GRDH_1SDV_20141123T224824_20141123T224849_003415_003FC6_6D14_ml.dim,/data/Sentinel-1A/work/20150615_111742/S1A_IW_GRDH_1SDV_20141205T224821_20141205T224846_003590_0043DB_C9D5_ml.dim,/data/Sentinel-1A/work/20150615_111742/S1A_IW_GRDH_1SDV_20150110T224823_20150110T224848_004115_004FC5_04F2_ml.dim,/data/Sentinel-1A/work/20150615_111742/S1A_IW_GRDH_1SDV_20150203T224828_20150203T224853_004465_005796_667F_ml.dim,/data/Sentinel-1A/work/20150615_111742/S1A_IW_GRDH_1SDV_20150215T224819_20150215T224844_004640_005B93_8C38_ml.dim,/data/Sentinel-1A/work/20150615_111742/S1A_IW_GRDH_1SDV_20150227T224813_20150227T224838_004815_005FD1_1603_ml.dim,/data/Sentinel-1A/work/20150615_111742/S1A_IW_GRDH_1SDV_20150311T224822_20150311T224847_004990_006408_BD29_ml.dim,/data/Sentinel-1A/work/20150615_111742/S1A_IW_GRDH_1SDV_20150323T224814_20150323T224839_005165_006832_5513_ml.dim,/data/Sentinel-1A/work/20150615_111742/S1A_IW_GRDH_1SDV_20150404T224822_20150404T224847_005340_006C58_9202_ml.dim,/data/Sentinel-1A/work/20150615_111742/S1A_IW_GRDH_1SDV_20150416T224815_20150416T224840_005515_0070B7_FADA_ml.dim,/data/Sentinel-1A/work/20150615_111742/S1A_IW_GRDH_1SDV_20150510T224816_20150510T224841_005865_0078CE_9C4A_ml.dim
-PoutputStack=/data/Sentinel-1A/work/20150615_111742/Stack_S1TBX_int.dim

I have 13 input scenes, some of which have slightly differing spatial extents. All the inputs to the co-registration process have been generated by performing calibration and multi-look processes on the downloaded S1A datasets. I am using the latest version of the S1TBX (v1.1.1) and NEST (v5.1).

Do you have any suggestions as to the nature of this issue or how I could modify the coregister.xml workflow to process correctly using the S1TBX?

Thanks in advance for your help and support.

Kind regards,
Steven