You would be entering in a very busy market with no obvious advantages. Let's look at the two types of LEDs that you might use:
If you're using inexpensive and low-power 5050 LEDs (15 lumens@350mW; usually ganged in groups), then you could power the lights off button cells (two CR2032s would power one 5050 LED for about 3 hours; 3 x 5050s for one hour) but even ganged you really don't have enough lumens to be compelling against $5-10 strap on lights.
If you're using high power Cree type LEDs (300 email@example.com watts), you can charge more but you are going to run into problems with battery life, installation, charging, and optics with your design.
I don't see sufficient depth/space in the design for sufficient optics. Again, not a problem for 5050s since they are so weak and diffuse. But for high-power Cree-types, you need beam forming reflectors and optics. The consumer won't be able to aim the light laterally with Crees (unless you have movable optics inside). The angle at which handlebars are bent viz a viz the stem is variable -- from straight-straight to a small bend to a severe bend. Not a problem for 5050s with diffuse lighting, but people at the higher price points of a Cree want a sharply focused light that goes in front of them.
Second, assuming you use Crees and 18650s, how are you going to charge them? For removable lights, you can just take them off and bring them in the house to charge. Many bikes are stored in sheds without power. Having to string a charger and micro-USB cable all the way to handlebars is going to be undoable for many.
Finally, regardless of Crees or 5050s, the positional relationship of the lights to the brake/shift levers is unclear and installation may be difficult. BMX trick bikes might not have brakes but most consumers will want lights on a street-legal bicycle. If your lights are in the grips, they'll be obscured by the brakes -- and your design is incompatible with grip shifters. If your lights are located further inward, inside of the brakes/shifters, then the consumer has to take the grips, brakes, and shifters off to install -- and you'd have to find a way to get your power (if using Crees) from the battery inside the bar to the lights. Installation would involve drilling the bar or tucking the wiring under the grip and past the brake/shifters. Not something that your average parent could do, so you'd have to add the cost of shop labor.
So because of these issues (optics, charging, battery life or installation difficulties) I don't see how you're going to be able to do it with Crees. You might be able to do it with 5050s but with the amount of diffuse light you're going to get out of it, I think your price point for 5050s is < $50 for a pair, which is going to be tough to make any profit for you.
Additional back of napkin calculations for those who wanted to know if it's possible with 18650s:
Using a single Cree XP-G driven at 1000 mA (250% luminous flux; about 300 raw lumens), you are burning 3.2 volts × 1 A = 3.2 watts. A high quality 18650 is 3.6 Ah × 3.7 v = 13.3 watt-hours. Boost driver efficiency at 85% gives you 11.3 watt-hours, so you have just under 4 hours of time with an 18650 at full (250%) power.
Mountain bikes have a handlebar diameter of 22mm. 18650s have a nominal diameter of 18mm (i.e., 18-650) but are often closer to 19mm actual because of the insulation and power protection circuitry). It's a tight squeeze. It won't work with many drops.
14500 lithiums would fit for sure as they're the same dimension as AAs (14mm D x 50mm L). However, they only have 900 mAh (fake ones will advertise more but are fake) which is 3.2 watt-hours or about 3 watt-hours after booster. So just about one hour of full (250%) power with one 14500, but you can serially gang the 14500 into two or three, depending on the curve of your handlebar (which would limit which bars it could fit into).