I asume this technology when mature enough will not be surrounded by a single point failure.

Up to a point, probably yes.

Allowing electric recharge of flying devices, that already have some battery autonomy, without having to land and take off, is clearly more efficient and less vulnerable.

Maybe is more efficient, but not less vulnerable. To recharge a flying device this way you basically mark the charging station even if you try to hide it, an attack could be carried against the station. Additionally having the drone or plane flying near it give away your position even if you come up with a mobile charging station (you cannot recharge too much far away, physic still stand). Then there is the problem of how much time you need to recharge it to a decent level, I am afraid that it would be in the hours range, and the necessity to keep the alignement, they had this problem also during the test, I suppose in a combat situation it would be way harder and this specific problem will not go away as the tech mature.

On the other hand, to keep the J37 Viggen example, it can be rearmed and refueled in 10 minutes and just need about 500 meters to take off. In this case if you don't see where the plane land, you also need the time to find it, it not give away its position during the operation with a microwave beacon.

Plus mobile electric rechargers, battery deposits and infrastructure will have the same weaknesses than fuel ones (except electric ones would may blow up a bit less under fire)

Once a battery is damaged, it make no difference that it blow up or not, it is useless. And generally a battery fire is harder to put out.

But it would be interesting to see how it eveolve and if it became mature enough to be used in a real combat situation.