Sounds like he was talking about the primary (first to receive fuel) to me. In any event, could the engine run (for more than a half hour) with any filter bowl half full of fuel?
Under the 'right' fluid conditions it 'is' possible.
The 'pleat pack' of pleated filter material flows from outside the pack to inside the pack and if the 'capillarity' of oil being 'held' in the open spaces (statistical 'pores') of the filter material isnt overcome by the motive vacuum of the system .... with AIR on the upstream/outer side being restrained by the filled 'pores' filled with oil, the oil with essentially no air still being 'sucked' though above the fluid level, then it is possible. A lot of 'serendipity' is going to have to combine to do this.
Usually, such a 'coarse' filter material will have a quite low 'bubble point' - the amount of air pressure needed to 'blow out' the fluid from the 'pores' and destroy the capillary action, but under the right conditions of partly dirty filter media, low temperature (higher viscosity) fluid, etc. etc. etc. ... it 'could happen', but not very 'often nor probable'.
That one observes air on the upstream side of a filter ... 99.99999% of the time will be caused by an upstream 'leak problem'.
Vacuum driven filter systems in most 'industrial strength' applications are usually strictly avoided because of their vulnerability to 'air leakage' (along with other very poor 'fluidic effects'). The very BEST way to filter for many many reasons is to use a pressure motive system. On recreational boats, a pressure feed system would require significant increase of component, manufacturing and assembly costs .... and the legal risk (the most important product in the USA, nowadays) to the manufacturer that could possibly cause severe fines by EPA/CG for a
design that 'could' wind up delivering small amounts of oil into the water is probably the reason one continues to see archaic and intrinsically flawed vacuum motive systems on recreational boats.