I would argue myself that it doesn't actually measure leakage currents at all, but rather estimates (and usually an over-estimation at that) what they might be!
No, it measures the leakage currents very accurately! But you have to understand the reason for the higher values than when using the Direct test method. Essentially the Alternative test is the same as performing an open neutral 60601 Direct leakage measurement, when it is normal to expect about twice the standard leakage current. Remember in a normal Direct test we have a T-NS mains supply in which the Neutral is referenced (connected) to Earth. Therefore generally half the capacitance that makes up the leakage currents have no potential across them, whereas when the Neutral is open the full mains supply is presented to all these capacitances. This is why both the open neutral 60601 tests and 62353 Alternative method allow twice the leakage values (1mA instead of 0.5mA).
The reason 62353 mentions you have to measure using another method if the measured Alternative method exceeds 1mA (or 2mA for mobile X-ray units) is this indicates the earth leakage leakage current may be at the 60601-1:2005 5mA level so this requires the touch leakage to be measured to ensure it is below 0.5mA.
From 62353: If the value of the alternative method exceeds 1 mA, either the direct method or measurement of the TOUCH CURRENT shall be applied (allowable values in Table 3).
One you switch to using the Alternative method and have reference values, as you need for all 62353 tests anyway, you can rest easy in the knowledge you have performed adequate safety tests to meet any legal obligations and saved time and money in the process!
As for your last point the choice of protective conductor test current is up to you, anywhere between 200mA and 25A. But it is generally accepted the lower figure is the one least likely to cause damage if an incorrect point is touched and therefore would be my choice every time. Better a false failure due to an oxidised contact than generating a potential fault by using 25A!
