I don't know test tool they were using but log from tool would...

I don't know test tool they were using but log from tool would have showed this instantly if they were using a modular tool that had multiple testers within perforated range otherwise they are guessing IMO. Rig on site - send MDT down with packers on string below and above and test again smaller sections individually or get more MDT modules on site measuring all sections sequentially on one string. Surprised as they would have been looking at drill cuttings and mud loss in carbonate structure along with traces of gas oil water and known high flows at lower level before casing if mud weight didn't just fill structure and huge mud loss at that level? Hope they can stimulate the top section they believe has shows and get a flow and that they have guessed some correct based on their assumptions. My assumption as a wild guess now is that they have misinterpreted mud loss as going into structure that was filled by gas but was infact filled with water - well at least the lower section and due to mud and mud weight used incase it was gas resevior. When flowing well test lower water flushed mud out lower section and top section not stimulated .
There is some hope and with workover rig onsite extra cost is minimal in big picture considering they know so much about structure. Hopefully other downhole tools help them determine exact structure they have.