ATK gets paid a little over $400m/year no matter how many SRB segments they produce. SRBs have additional costs in assembly, integration and logistsics, but USA and NASA folk are paid for that. ATK probably doesn't care wether the stick or stumpy is chosen, but the USA and NASA guys see twice as many SRBs to assemble (most of the work is on the tip and tail) that might keep them as busy as they are with shuttle, which is politically convenient.<br /><br />Stumpy has one huge, glaring disadvantage over the stick - the 3 seg SRB will have a different casing, nozzle and grain geometry than the 5 seg. This will require a big dollop of R&D, though perhaps less than the 5 seg - but in addition to the 5 seg, and ATK, USA and NASA will require additional cash to carry two different types of SRB at the same time. <br /><br />Stumpy's core stage does closely resemble the EDS, at least in engine configuration and diameter, so it may save a modest amount of work on that part of the HLV. It's possible the structural strength needed to hold onto the SRBs might be close to that needed to support 130t of payload, in which case the EDS and stumpy core could be almost identical.