"Biasing" in this context basically means: to provide power to the active antenna
Active GPS antennas have a built-in amplifier (you'll see the term LNA floating around, Low Noise Amplifier), near the actual antenna element. The signals from GPS are so small that losses from a long (~10cm or more) wire are incredibly detrimental to the signal.
For any amplifier to work, it needs power (a positive and a return, aka ground). Most GPS connections use SMA connectors, and they only have 2 pins. One for ground, and one for "data". So how do you "send power" to the amplifier that is built into the active GPS receiver when we don't have an extra power pin? This is where the inductor comes in - it's a trick to share the positive power line with the GPS "data" signal line.
This works because the GPS data is completely AC data. The inductor decouples the output signal from the high impedance positive power rail, allowing the output signal's AC component (the actual GPS RF data) to come back along the power line. Or another way to say it, prevents the power rail from washing out the signal that we care about.
And yes, that extra 22pF capacitor is just to help provide clean (low noise) power to the inductor. If there is extra noise on the power line, that noise can get through the inductor and wash out the GPS signal that you care about.
No, you need to leave a hole through the solder mask under the inner feed and provide an isolated pad to solder it to. Anyway, if you feed from the inner feed, then the outer feed is going to become a stub that could cause some rather major reflection issues.
Best Answer
GPS (or any radio) RF trace to the antenna should not go across the whole board because of the interference, as it appears in your design. Some common sense suggestions are:
In my opinion you should either use the SparkFun board that is properly tuned and tested, or transfer it completely, including the connector, and choose a PCB material that is similar if not exactly like the SparkFun board.