The most obvious application for this would be to stick a camera on it.
Launch it. Then have it in the air for the duration of an entire day, making small manoeuvres to get best camera angles and footage, if required.
Using GPS locking, the blimp should be able to hold a position, even in windy conditions.
This could be good for sports entertainment, in stadiums or even for security purposes during big events.
Where a quad copter would run out of battery in about 25-30 mins, this object should be able to stay in the air for a whole day and stream live images to the control centre.
For me, your best bet is to look at ArduPilot's tricopter implementation. That is going to be the most likely software which will run on your blimp, out of the box (though some modifications to the physical mounting of your motors will be required). To do this you may need to remove the tilting mechanisms you have created (I know it sucks, right) and place the motors in the configuration which is identical to that of a tricopter. This would give you the manoeuvrability of a tricopter, though with the added flight time of a blimp. You will need to create some software, which completely cuts out the motors, when the blimp is successfully holding its position (no need to spin the motors to stay in a position, because the blimp is keeping the object in the air, not the motor/props).
I actually don't know why other people haven't done this yet, because for filming static locations, the blimp idea completely solves the problem of drones being unable to stay in the air for long periods of time. The disadvantage is that this object will be slower in the air, though this wouldn't be a problem if all you need is a static camera position. You can always move the camera (gimbal) independently of the blimp, to ensure the camera is always pointing in the correct position.
BTW, how much weight can your prototype blimp handle?
PS. I understand that you want to write your own code, so you have absolute control over it, etc (I too have done this in the past). But why not use code which is already out there and has been tested by 1000s of people? I would use ArduPilot and modify it, to suit your requirements. This will save you a lot of time, IMO.