Question 1: About 0.95.

Question 2: Ask people who knew me? Infer a model of my mind from that and my writings? I don't consider it more ethical to use uncertainty as a reason postpone it until some unforeseeable technology is developed.

Question 3: I'm reluctant to enter such a lottery because I don't trust someone who believes those assumptions. I expect the scanning part of the process to improve (without depending on human trials) to the point where enough information is preserved to make a >0.99 fidelity upload theoretically possible. I would accept a trial which took that information and experimented with a simulation of 0.5 fidelity in an attempt to improve the simulation software, assuming the raw information would later be used to produce a better upload.