The loss from the broadband matching device is twice the loss from the tunable matching device in my tests. The loss is low, only about one DB (vs a half DB) if the SWR is low. IF the SWR is above 1.5:1 the loss in the broadband matching device goes up exponentially. You will find with this broadband matching device the SWR never seems to go above about 4:1 because it is acting like a dummy load. Variation in the environment such as wet vs dry ground, rock vs soil, trees vs none, different heights, different poles, can cause the wire to no longer be exactly tuned the way it was in the back yard.
Loss with RG174 can be significant at frequencies above ten MHZ. If you start to add losses while running very low power, it can mean the difference between QSO and no QSO. If you use a feedline, I don’t, better coax that is short isn’t going to weigh any more and so might as well be used.
I don’t use a feedline at all. An endfed HALFWAVE has essentially no current, no radiation at the end of the wire. I simply use four feet of the EFHW as my feedline.
All these antennas and the matching devices can be made yourself at a fraction of the cost. Making and tuning them is educational.
73 fred kt5x / ws0ta