Obviously, for high frequency (eg PWM) switching you want to reduce the pull-up resistor, or even better use a push-pull. However, I can't see the OP going for a push-pull since he seemed convinced that even one BJT in the circuit was one BJT too many.
The circuit I posted is ideal for low speed switching - things like turning a light on and off, or removing power from a subcircuit in a system for power saving or hard-reset facilities. Things where you're not going to be switching it on and off faster than you can blink 
Also, I tend to use a MOSFET that is an order of magnitude greater than I need as regards current handling ability - my most common one is a 3.5A one (SOT-23 even at that current), and I rarely switch more than a few hundred mA per MOSFET. Maybe controlling the power to a radio transceiver module, or a display, or similar, to save battery power. I have one project here which is a solar powered remote temperature monitor - samples the temperature every 15 seconds and relays it to a base station via ISM band. Runs off one LiPo battery (stolen from a phone). Charges up from flat in a day in good sunlight. Will run for a couple of months with the solar panels disconnected - and all because of that switching circuit.