The argument is that although there is neuronal activity, it is an activity so degraded that according to all we know it could not cause lucid and coherent experiences as NDEs.
Yes although there is some evidence of activity that can reach the cortex this is not to suggest that the activity caused by quasi-rhythmic waves emanating from the hippocampus could possibly produce the highly lucid and complex activity we find in an NDE.
Also the study seems to suggest this is not something that can occur straight away when someone enters a flat-brain state (at least as far as cortex activity is concerned) as granule cells prevent ripples spreading until a threshold of activity is achieved (at least that's how I'm reading the bit quoted below):
"Therefore, marginally synchronized ripple events will not be relayed to the cortex by the EC but will be directed towards the hippocampal DG via the perforant pathway. As mentioned above, DG granule cells show very low excitability and have been described as “gate” or “filter” for hippocampal afferent signals
[35],
[36]. Thus the relatively unsynchronized EPSPs of ripples are filtered out by granule cells and their propagation within the hippocampus stops at the DG gate (
Fig. 8A). However, since the intra-hippocampal synchrony steadily increases in successive ripple events, the continuous excitation of the granule cell gate by ever increasing volleys leads to a decreasing activation threshold, as granule cells appear to be highly sensitive to preceding signaling activity
[37]. Also, CA3 synchronization may have to undergo a certain threshold behavior before single-cell bursting activity leads to a truly synchronous population burst
[38]. When this threshold is reached, DG activation is achieved. The divergent connectivity of granule cells
[39] ensures their synchronous activation, which in turn results in highly concurrent propagation of the signal towards the rest of the hippocampus via the CA3: a ν-complex is generated (
Fig. 8B). This event is then carried through CA2, CA1 and subiculum further towards the EC, where it now activates the deep V–VI layers instead of superficial layers due to the νC’s high frequency signal. Consequently, the νC event is further directed towards neocortical targets (
Fig. 8C) from where it spreads into virtually the whole brain"
Its unclear the timescale involved here - however this may also be an issue in using this data to explain ndes as ndes can occur very quickly following flatline in cardiac arrest.
But hey I'm not a neuroscientist so feel free to shoot me down!
D