Display analytics reports what the network did: which asset played, on which screen, when, for how long, and whether the device was actually awake. In advertising this is proof of play, and it is what an invoice is defensible on.
It is also the only honest basis for content decisions. A campaign that underperformed because it was poor and one that underperformed because its screens were offline for a fortnight look identical in sales data and completely different here.
The caveat is that proof of play proves playback, not viewing. A screen can faithfully report ten thousand plays into an empty corridor. Combining play records with any measure of presence (footfall, dwell, interactions) is what turns activity into evidence, and platforms that report only the former should be described as reporting uptime.
Ask specifically what is logged when a device is offline. A player that continues playing locally and syncs its log afterwards gives you the true record; one that logs only at play time reports a gap that never happened, or worse, claims plays it never made.