though it is perhaps diffivult to get tight bounds in open cells.
The simple Two Zone Model applies in either case. Here are the primary differences:
In a closed cell, in order to avoid a pressure rupture (which would be most likely followed by a flash fire) there is a recombination catalyst in the gas space of the cell. This is missing in an open cell. Instead we have a vent port through which electrolysis gases, water vapor, and entrained liquid water/electrolyte exit the cell.
Both types of cell typically (not always) use basic electrolytes. This always dissolves a little bit of the Pt and Pd, which then can deposit in other places. This has been shown to be on the electrodes by surface studies of used electrodes. The dissolved metals, which are in ionic form, can also be splashed onto the sides of the cell in the gas space. There they can (not always) be metalized by reduction with hydrogen, which forms a 'nano'-sized metal particulate on the wall that is an excellent catalyst for recombination. In the closed cell, this will be very difficult to detect since the recombination at the metal particle has not moved to a different enough heat capture efficiency zone. However, in an open cell it should be easily detectable (given the usual detection limit considerations) by two methods: Loss of lost water volume and detection of the recombination heat. Said heat will be in the lower efficiency zone of the Two Zone model, but in fact it may well have another associated detection efficiency, since it is well removed from the major loss pathways represented by the penetrations through the cell top, so it may well be under measured. Details matter here,and the Two Zone Model is not sophisticated enough to quantify this well.
When ATER begins in an open cell, the recombination heat would be measured no differently from a closed cell, i.e. it is showing up in the high efficiency zone and will be inflated by 'old' (pre-ATER) calibration, giving an inaccurate apparent excess heat.
But I don't see any particular reason to believe the open cell case would be any less accurate than the closed cell if calibrated as I suggested with two heaters in the cell, one in the liquid and one in the gas. Always open to valid corrections of course.