I understand that the P-175 polecat has two engines which normally would mean one air intake for each engine, however other aircraft have used 1 air intake structure for two separate engines such as the Eurofighter Typhoon and the B-2. Furthermore it seems that most flying wing UAVs these days tend to have one centreline air inlet. Wouldn't it have saved some complexity to have both engines fed by a single air intake?
As other answers have indicated, there are a lot of complex tradeoffs involved -- the details of which may be classified and/or proprietary. Those who know the answer can't answer -- the rest of us can just guess.
When designing an aircraft for low observability (in this case radar), you need to shield the fan blades from visibility from outside the aircraft. Consequently, LO aircraft end up burying the engines and using S-duct or covered inlets.
For a single-engine aircraft, this may place the inlet on the center line like an X-47B. In order to hide an inlet, it seems you need an S-duct with an offset of about one engine diameter. For a centerline installation, this means your aircraft thickness at the center line is at least two engine diameters.
Turning the flow a lot in a short distance can cause separation from the duct walls. A large amount of turning makes for a long duct -- more duct losses -- more weight -- longer aircraft, etc. Nobody wants a long duct.
By splitting the inlet in two and laying each S-duct on its side, the offset may be reduced to one engine radius (each). This may make for a shorter duct. The maximum overall thickness may be kept to one engine diameter.
This also opens up some centerline real estate on the aircraft. Something can now go between those inlets. It may not be important in this case, but sometimes locating satellite dish, cameras, other sensors, lasers, radar, accelerometers, gyros, etc. on the centerline can have advantages. At a minimum, weight and balance prefers a symmetrical aircraft so anything that you only have one of is most easily placed on the center line. Famously, the A-10 valued a centerline location of the gun so much that the nose gear is offset.
Air inlets design is normally a straightforward matter.
It has to provide the engine(s) with a clean, sufficient and relative slow airflow. In particular, for a standard turbofan or turbojet, airflow speed should be reduced to some Mach 0.4/0.5 which is the optimal value for the engine's first-stage compressor. And as a rule of thumb, for each 1% of pressure lost in the inlet, the engine loses some 1.5% thrust. That's why an inlet is basically a pipe with a gently increasing section which, according to Bernoulli, slows the airspeed down and increases pressure.
Furthermore, inlets are normally a separate structure sticking out of the fuselage therefore adding additional weight and drag to the aircraft. For these reasons, inlets should be as few, short, straight and compact as possible.
Their positioning depends on the airplane's mission. For an aircraft like the B-2, the F-117 and the P-175, engines are mounted inside the wing both to reduce radar and infrared traceability and to have a free fuselage for the payload. For the same reason of reducing traceability, inlets and outlets are mounted above the wing. Having one engine (or one couple of engines) on each side of the wing, one single straight and short intake for each side of the wing is therefore the best solution in these cases.
On the contrary, having for example the RQ-170 Sentinel one single engine, it is better served with only one intake over the nose.