Why are most RGB LED strips common anode instead of common cathode?
The reason common anode is more common is because its easier to sink current than to source it. With either common anode or common cathode you'll have one terminal connected directly to a supply for all LEDs and the other side having the dropper resistor and a control transistor per pin (or IC outputs that are transistors on the inside) either sinking or sourcing a current.
NMOS / NPN transistors are stronger in general, more common as discrete and are better at sinking current than sourcing. You need PMOS/PNP transistors to source current (pull up) effectively, but they'll still be weaker at sourcing than an equivalent N-transistor would be at sinking. Thus the best solution is to connect a common anode to the positive supply and sink current from each LED using NMOS transistors.
Older ICs used to be designed exclusively using N transistors for speed reasons, and so were much better at sinking current than sourcing it. This was particularly true of the TTL logic used in the 74LS series chips (still widely used as interface chips). A 74LS00 is specced to sink 4-8mA, but source only 0.4mA.
Modern CMOS ICs are much more symmetrical (an ATMEGA328 in an Arduino can source or sink 20mA) since they use bigger PMOS than NMOS to balance the fundamental differences, but the convention of common anode is well established.
EDIT (More info): If on the other hand you're building a matrix, you'll have to have both current source and sink transistors. In this case it can be best to have more devices in common cathode and fewer on a common anode. The idea here is to have a few fat NMOS devices sinking many LEDs currents and many weak sources (I/O pins) driving a few LEDs each. Of course with common anode strips you could use fat PMOS devices too.
Other answers have mentioned that N-channel transistors are cheaper than P-channel for the same current capacity, which is a reason to put the switching transistors on the negative side of the LEDs.
Another reason has to do with thermal management. Most LEDs are grown on a chip with the cathode layer underneath and the anode on top. The chip is thermally bonded to the pin it sits on, but you have to let the light get out, so the anode is just fed with a tiny wire coming over the top of the chip.
This means that if you're putting red, green, and blue chips in one package, having multiple cathode pins and a common anode will give better heat dissipation -- each chip will have its own pin to drain heat out of, going into its own trace on the strip.
I can suggest a couple of reasons why common anode are favored:
Safer wiring. A wire which completes the circuit for a remote device often must travel some distance through mechanically stressful conditions. It's preferable for that wire to be at Ground voltage rather than plus supply voltage, so that if it shorts to the chassis or other wires there's less hazard.
This, in combination with customary use of positive-voltage power supply rather than negative, leads to favoring separate cathodes for LEDs.
NPN transistors easier to manufacture than PNP. NPN transistors (in silicon) have had a better price/performance ratio than PNP transistors, as explained by this random article here: [Why NPN transistors are preferred over PNP?] (http://www.madsci.org/posts/archives/2003-05/1051807147.Ph.r.html). It is the switching and amplification configurations possible with each kind of BJT that is part of what motivated the preference for positive supply voltages.
And for switching purposes, a BJT transistor needs to be used in common emitter configuration, which, for NPN used with positive supply, means switching the low-side (cathode) side of the LED.
In my experience it is easier to switch the negative side.
Lots of electronics will have different voltage requirements. When you connect a lot together (for example, an LED or LED strip and a micro controller) they will likely have a common ground, but different supply voltage. Most voltage regulators will have a common ground, a high voltage input and a low voltage output.
To switch the cathode (or ground or 0 V side) you can use a logic level, n-channel MOSFET. This will require the gate to go to a few volts above 0 V for the transistor to be on and 0 V for it to be off. This is typically quite easy for micro controllers which go to 3.3 or 5 V.
To switch the anode (or positive side), for a device operating at a higher voltage (say 12 V), you would use a logic level p-channel MOSFET. This requires you to supply it a range of 0 V to a few volts below the supply level (12 V). This means that a 3.3 V or 5 V micro controller is unable to directly control the transistor. Instead you need to add additional devices such as an n-channel MOSFET and a few resistors or an opto-isolator and a few resistors and so on. The other option would be to have a common positive voltage of 0V and have the negative voltages be negative (so -3.3 or -5 V for a micro controller and -12 V for the LEDs), but this then requires you to make sure the negative voltages aren't connected directly, which would prevent say running a LED light strip and an Arduino off the same power supply without other additional electronics.
As such, switching the cathode is typically much easier.
Because you want to control the colours individually, that makes having a common anode (and thus individual cathodes) an easier way to switch them.
I wasn't able to find any definitive reasons, but I did come across:
It has always been my contention and design technique to sink current where possible, rather than source it, thus I prefer common anode wherever possible for displays and other driven devices and I write all firmware routines to provide lows for execution rather than highs.. The reasons are obvious on most data sheets that most devices can sink more than they can source.
— EEng (source)
It may be that the slight advantage that sinking current offers over sourcing for most devices, leads manufacturers more often to design displays in a common anode configuration.
Probably like anything else, the invisible hands of the free market moved both manufacturers and consumers alike to common anode simply because more people bought common anode. Almost like Darwin's theory of origin of species. Two animals cannot occupy the same niche, one will dominate the other. Why did AC win over DC? Why did VHS win over Betamax? Generic Flash MP3 players vs Zune vs iPods? Because one was favored over the other, and manufacturers followed suit.
LED strips are different from normal electronics parts, because there is much direct enduser and consumer purchasing. And the mass production manufacturers who copied the initial offerings will only mass produce what's profitable.
Manufacturers see Consumers purchasing Common Anode, so they produce more. Consumers see more Common Anode, they purchase more. Chicken or Egg, the end result is the same.
My guess is that this practice may go back to days when using TTL 7405 chips was common (early MOS chips being too weak or far more expensive). The open collector 7405 (et.al.) could only pull down, and perhaps pull down against something other than a regulated 5 Volts connected to the Anode of an LED.