Fix error handling in dbt build

[jtcohen6]

resolves #3598

Description

I saw that the BuildTask was inheriting from CompileTask, which sets:
https:/dbt-labs/dbt/blob/9f716b31b3e0ca57a3722809f8c3405a7b5752f8/core/dbt/task/compile.py#L36-L38

That's the cause of the weird behavior I was seeing when reviewing #3490. We could just override that method for the BuildTask, but what I think we really want is for dbt build to operate at parity with dbt run, and take all of its cues—including how it handles hooks, deferral, etc, etc. I confirmed that TestTask, SeedTask, SnapshotTask also inherit from RunTask, so it feels appropriate that BuildTask would as well. Let me know if I'm totally off base here.

Tiny thing: I confirmed that a failed model will be handled and downstream resources will be skipped. Now that tests can be skipped, in addition to models/snapshots, I figured we should tweak the SKIP log message to be more inclusive of tests.

Checklist

• I have signed the CLA
• I have run this code in development and it appears to resolve the stated issue
• This PR includes tests, or tests are not required/relevant for this PR
• I have updated the CHANGELOG.md and added information about my change to the "dbt next" section.

[iknox-fa]

Good catch! I think I chose to inherit from CompileTask ealy on when I thought I needed the selection spec stuff found there.

[jtcohen6]

Some failing tests, but they all look related to the flakey issues we're currently tackling elsewhere, and unrelated to the changes in this PR.