For all three of your questions, the answer is the same:
Technically, Nothing
A hacker's master key ability doesn't help him create codes at all. Decreasing a code's Power Rating doesn't do anything to change the DCs of the hacker's Science (Mathematics) skill check needed to create a code.
It's a hole in the rules--an annoying one, given how tightly written Spycraft 2.0 is.
What It Should Do
I believe the final phrase in the master key ability ("...for the purposes of determining skill and Request check DCs") is legacy text leftover from concerns about Power Ratings stemming from the gear chapter having been--if I'm remembering Spycraft 2.0's development correctly--written independently of the remainder of the text.
Excising that text (and house ruling in a (minimum 1) in the same section after reducing the Power Rating) allows the master key ability to be useful when creating codes by, potentially, vastly reducing the amount of time necessary to create codes and software because of the check's PJI (High) tag. This drops an increasingly powerful code's corresponding Complexity and permits the hacker to create codes faster.1
I've compiled the chart below so you can see the impact this would have.
Code's Code's Complexity: Minimum Project
Power Rating (2 × the mission's Threat Level) + Investment Per Check
1 13 1 week
2 15 1d3 weeks
4 18 1d3 weeks
6 20 2d4 weeks
8 23 2d4 weeks
9 or 10 25 3d4 weeks
This time is spent per challenge the character undertakes (p. 101) with his Science (Mathematics) skill, so dropping the Power Rating even by 1 can save weeks or months of work.
- I've never had a hacker create a code beyond PR 1. Almost all the skill checks with the PJI (High) tag are ridiculous, given the tiny amount of time available during most missions. Cash can be gained by ruthlessly exploiting the rules (e.g. earn millions in minutes by putting a UAV on Ebay), but time? That's a hot commodity.