A bunch of different topics brought up so far. I'd like to discuss all of them.
The draw from your gate drive is going to be insignificant compared to the draw of an atomizer. So, whatever the value of your pull up/down, it's not going to have any noticeable effect on battery life. It's really a non-issue in that regard. The main thing you want to concern yourself with is any current draw while the device is inactive. A full time current draw of even a few mA will have a noticeable impact on run time.
For a simple mod circuit, there normally would not be any idle draw on the battery since the switch opens the circuit completely. The leakage of the MOSFET gate is a barely measurable current. If you bypass the switch with other electronics, then you have a different situation. That's not typically the case for a basic atomizer circuit, but for more complex circuits such as those that utilize a micro-controller, idle current draw becomes a major concern.
Several of the N-channel touch switch diagrams I've seen here show a 1k resistor in series with the touch pad and pull down resistor. I don't really seen any function for that resistor. It's going to have zero effect as ESD protection and is redundant as far as operation of the switch. I don't understand why it's located there. To protect the positive pad from shorts for an N-channel touch switch, the 1k resistor needs to be between the the battery and the pad. For a P-channel touch switch, there's no reason to add any resistor other than a pull up.
Employing a PTC fuse allows you to safely run an unprotected IMR cell. Some people are under the mistaken impression that IMR cells are perfectly safe. They are not. They can vent the same as a regular ICO cell. The difference is that it takes a much more severe condition for that to happen.
Ideally, an IMR cell should be protected from short circuits. A PTC fuse is a quick and easy way to do that. Atomizer and cartomizer shorts are actually not that uncommon. I've had a couple myself. You want the fuse as close to battery positive or negative as practical to protect it from as much of the wiring as possible. You might as well cover wiring faults as well. Positive or negative doesn't matter, but conventionally, the fuse would be located on the positive battery terminal. Personally, I feel the presence of a fuse negates the need for a master switch, but that's also a matter of preference.
MOSFETs have an amount of capacitance at the gate terminal. Since it varies with current flow, the value is given as gate charge instead of capacitance in the data sheet. It can be a primary consideration for some applications. Since the effects of capacitance become more pronounced with higher resistance, it's been mentioned when using the very high value resistors required for a touch switch.
Gate capacitance dampens switching response making it look like a ramped up and ramped down delay. For a switch that has to turn on and off quickly, gate capacitance can slow down response enough to make a circuit non-functional. However, we're talking about reductions in the hundredths of a second. Maybe a tenth of a second for a worst case. I really don't think that is going to be an issue for turning an atomizer off and on with a touch switch.
Ringing is another consideration that is related to gate capacitance, but doesn't really apply here. On a scope, circuit ringing looks like the sound waves of a ringing bell, hence the appropriate name. It's a type of response that has to do with a circuit's resonant frequency and is characterized by a combination of resistance, inductance, and capacitance. All circuits have some amount of resonance, even if very small. Ringing can become an issue even at fairly low frequencies. In the case of a touch switch, it's not going to come into play. When limited in consideration to the step input of a manual switch, there's not enough inductance in the circuit for ringing to become an issue.
Besides, I may be wrong, but I think it may limit the Source - Gate current, improving battery life (though probably not very much). Your whole goal is to pull your gate back to B- to unlatch the FET.
