Quigsworth, easy to test your suggestion: just connect a LM2596 board to batteries and measure input current. I don't think that it will be close to zero. If so, also imagine that something got wrong with the board (bad capacitor etc). In this case if you haven't an input switch, current consumption may be out of your control, so explosions, bloody fingers etc. I don't think that your recommendation is suitable for electronics novices.
I like it

, constructive, provocative...makes me go hmmm...(I love this forum

)...and Slimest, your point of not recommending this switching method to novices is well taken...
...I read your response last night and decided to think about it some more...one thing I didn't do was qualify my suggestion by taking into consideration the deployment of this reg (a pretty big oversight on my part

)...is it a PT, paired with a PSU?...is it hard-wired into your car like I have done?...is it in a hand held device?...is that hand held device also a PT?...does it have a meter?...and to throw a "real" big curve into this discussion, is it being tact switched? (but let's leave that one alone, it's a game changer)...the list of uses goes on and on with these regs...
For the sake of keeping this meaningful, let's go with Foxman's use. It's a great example of taking the basic batt/switch/atty box to the next level.
1st off, I am by no means recommending that there "not" be a master on/off switch between the batts and reg, that should be a given but it should be noted, when you're done vaping and you put the device in your pocket/bag/whatever, turn it off (or pull the batts)....that being said,
There are inherent risks to these devices we make and having a constant potential on the input and the issues that could arise from that has crossed my mind on a number of occasions...I'm forever forgetting to turn off my shop laptop PSU PT (I've now put an LED behind the master on/off

)...but after weighing the risk pro's and con's I still came to the conclusion that
I would rather be holding a reg that is warm with a stand-by current flowing through it when I press the fire button than one that is cold and with a button press, instantly connect 7.4v and a capacity to run 10A+ through it...in
my experience, that's when components "pop".
...is there any empirical evidence that suggests that one method is safer than the other? I guess 2 regs could be bench tested to destruction with a timer and a relay, might be worth it "just to see"
As a final note, I don't really take issue with seeing either switching method, both have their pros/cons but if I ever post a handheld device with an LM2596 switched at the output (and it will be) and someone chimes in with "you've done it wrong", be prepared to back-up your statement with a clear, cogent explanation as to "why" you think I've done it wrong
