It's basically because they're doing it wrong, and to a point you're argumenting wrong.
What you need to do is find 3-5 targets and cycle them, thus not using value on scans every time. It's important that these targets are inactive and holding a reasonable pool of resources.
Trust me they exist.
now all you need to aim for is hacking 3000 of each resource, (2500 for the mission cost, and 500 each for - over time, covering the hiring, the initial scan for finding the planets
(those should be few no more than 10 scans of each, if you know what you're looking for, using tools)
As the game runs these days, you can actually get it free on #scans on irc
the odd fail (allthough those should be few, as your target is an inactive git,, so you can predict the alert fot every mission)
Let's say that amounts to 100000 of each resource over the lenght of a round that would give you room for 30 failed missions + some, you shouldn't have more fails, or you're doing it wrong.
So over a rounds lenght 1177 ticks you need with the above example, 200 succesful missions in 230 attempts to break even.
As a cath you can manage a few more than 230 over a round 
Now this is a theoretical discussion because few are willing to sacrifice research points at an early stage in the round, especially those going for top 10 or 50.
If the bank hack was 50 research ticks faster done on average, I think it would be viable for most if not all cath to go fast bank hack.
Again that would mean 100+ planets suddenly back hacking away, increasing the failure chance.
So yes and no