Yup, it will definitely work. You could even throw together two quick (preferably Schottky) diodes and say... Two 100nF ceramic caps, (like the setup I made as a proof of concept for this answer) instead of an IC charge pump. Anyway, before I get too out of scope...
An LCD's Bias Voltage affects the contrast of the backlight (background) to characters (foreground), as well as the available viewing angle. Normally, a potentiometer's wiper is connected to Vo; this is the LCD's V Bias Input. The potentiometer -normally- acts as a voltage divider, between Vdd and Vss. Like I mentioned, -normally- at +5V Vdd and 0V Vss; this is important because of your datasheet!
For example, if you can find a Shenzhen datasheet floating around the web (as it's likely you have a ubiquitous Chinese LCD), you'
ll find something like this:
Operating Voltage for LCD: Vdd-Vo Typical 5.0V
Now, this means that with our hypothetical 'normal' situation, we have Vdd at +5V, Vss at 0V, and our potentiometer divides that into something in between - thus providing Vo a voltage which affects our contrast and view angle!
So, how does running one of these puppies at +3V3 work? Well, I'm sure glad you asked. Remember that datasheet? It says we ideally want our operating range to equal about 5 volts, but of course there's a bit of lee-way there. Different applications require different contrast and view angles because... get this.. People's eyes are different!
Now, if we powered the LCD from +3V3 and kept our Vo referenced to a 0V Vss (Ground), we wouldn't have enough voltage bias on Vo to allow us to see. (At least not very well...)
Vdd = +3v3
Vss = 0v
Potentiometer connects Vdd on one leg to Vss on the other leg. Vo is connected to the wiper.
Our LCD's potential range is now only 3v3... Not enough to see much, if anything!
What if we use a negative voltage bias?
Vdd = +3v3
Vss = 0V (still)
Potentiometer connects Vdd on one leg to our Negative bias on the other leg. Vo is connected to the wiper.
Negative bias = -1.6V
3v3 minus -1.6 = 4.9v!
Wowee! Suddenly Vo has nearly a full 5v potential range! Now our potentiometer can properly divide this ~5v range into something our eyes can see! With light!