dashboard/app: more prioritization signals

[tarasmadan]

Is your feature request related to a problem? Please describe.
Some customers want to see alternative signals to prioritize their bug fixing efforts.

Plan

• Sort the existing reports by projected impact.
  • dashboard/app: sort bugs by the impact by default #6071
• If there is some stronger signal under the detected one - show it.
  • syz-manager: add Rank column and show the bug consequences as a tooltip #6271
  • dashboard/app: integrate tail reports #6332
  • dashboard: reconfigure syzbot to not panic on WARN
• What are the minimal required credentials to trigger the bug?
  • ? TODO: decide whether we want to scale the config variation first. Requires discussion with @dvyukov and @a-nogikh .
  • ? TODO: either teach existing instances to drop credentials OR run more specialized instances
• Bust the new prioritization signals. The quality of the signals should be maximal now.
  • ? do we want to order by Rank by default?
  • dashboard/app: make sorting bugs by Rang the new default
  • TODO: add tags
  • TODO: shutdown bad-writes instance
  • What to send additionally in e-mails?

[tarasmadan]

@a-nogikh believes it is better to not touch the default ordering #6271 (comment) .
@dvyukov thinks it makes sense to order by Rank by default #6071 (comment) .

The proposal is to not touch it now and reconsider at the end of the Ranking story (this PR).

[tarasmadan]

Discussed the "decide whether we want to scale the config variation first" with @ramosian-glider @dvyukov @a-nogikh offline.
The complexity of the alternatives looks too high. Let's land a few specialized instances instead.
The specific candidates are:

1. Non-root KASAN manager. What repos?
2. Next potential candidate is the KASAN instance with CAP_NET_ADMIN.

Overall it looks promising to set "all sanitizers" for "all trees" on "all architecture" with "different credentials" as a goal.
This is probably the ultimate (not realistic) goal for our infrastructure.