The current analysis of coordination in the Matrix seems to be attempting to use both R/L-INDEX and R/L-HNDL. Not sure if this is really desirable, but the current status is also half-baked: in VP/S coordination, the L-HNDL gets identified, but the HOOK information is lost in vp/s-bottom-coord phrase. (The mystery is why this is different in the mmt/eng grammar.) Adding this seems to fix it, but I'm not 100% sure that's the direction we want to go.
The current analysis of coordination in the Matrix seems to be attempting to use both R/L-INDEX and R/L-HNDL. Not sure if this is really desirable, but the current status is also half-baked: in VP/S coordination, the L-HNDL gets identified, but the HOOK information is lost in vp/s-bottom-coord phrase. (The mystery is why this is different in the mmt/eng grammar.) Adding this seems to fix it, but I'm not 100% sure that's the direction we want to go.