Coverage metric in report

[therealryan]

We add masking operations to messages to deal with message content that we don't want to assert on in testing (it might be unpredictable or just boring to model). These operations necessarily reduce test strength, as we're no longer asserting on the entire message.
Currently the only way to see how much of the message is actually being asserted on is to go to the execution report and manually compare the full and asserted messages. This means it's possible to add over-enthusiastic masking operations that hide message content that we really should be asserting on, and it'd take someone doing a bunch of clicking around to spot the weakened coverage.

It'd be good to highlight test strength in the execution report to reduce review effort.

For each asserted message in the report detail page, we should compare the asserted (masks applied) and full (no masking) messages and compute a test strength metric, ranging from 100% (the asserted and full messages are exactly the same) to 0% (the asserted and full messages are completely different) and display that alongside the assertion result.