Measure 0-RTT amplification

[mpiraux]

Kind of the same test as Address validation, but using a 0-RTT packet and measuring the total amplification.

CC @rmarx for the idea

[rmarx]

So what I'm doing is also measuring if the amplification prevention is dynamic. Practically speaking, the spec says 3*received_data, so I also fluctuate the size of my 0-RTT data (by creating large HTTP headers atm) to see if the server just has a fixed limit (e.g., 3*1280) or actually looks at how much 0-RTT data is received in practice (and then, how that interacts with the init cwnd as well).