The issue is a significant hardening of the pT-spectrum after showering in POWHEG+Pythia8 for certain observables, even for non-zero values of the POWHEG damping factor (specific examples are gg→HH, see arXiv:1703.09252, or gg→H, see arXiv:1510.08850).

Here are a few questions/topics that arose in the discussion:

• Can one identify classes of observables where the effect is more prominent than in others?
• Is the effect specific to Pythia8, or is it observed also with Pythia6?

Here are example plots hdamp=250, hdamp=infty from gg to HH which suggest that the effect is more pronounced with Pythia8 than with Pythia6. The plots are also in version 2 (journal version) of arXiv:1703.09252.

• Is the effect specific to POWHEG?
• In arXiv:1510.08850, it was proposed to restrict the shower scale in the remnant events: is this legitimate for all observables?