`debug zip`: put table dump output rows on a single line

**Is your feature request related to a problem? Please describe.**
TSEs have brought to our attention that table dumps contained within `debug zip` in the default `--format` can span multiple lines in tables such as `crdb_internal.jobs`.

This makes existing workflows involving `grep` for TSE difficult. Ideally, with `grep`, all the information is contained on a single line within the file.

**Describe the solution you'd like**
Come up with a way to ensure that table dumps contained within `debug zip` maintain a "1 line per row" format **in the default case**. Something like find/replacing newlines, such as in https://github.com/cockroachdb/cockroach/pull/111003, are being considered, but require more careful thought (for example, if you replace newlines with `\n`, how do you differentiate from a legitimate `\n`?).

**Describe alternatives you've considered**
TSEs are our primary stakeholders for `debug zip` and we need to meet them where they are. Despite 3rd party tools being available that could help here, we need to understand where TSE is coming from. Many times, TSEs are accessing debug zips in different environments, many times constrained regarding what tools they have access to. If TSE indicates to us that something needs to work via the use of `grep`, we need to satisfy that need.

Also remember that this needs to be **default behavior**. Every time we introduce a new CLI flag that's not default, that adds toil and confusion to the debug/escalation process. Many times introducing changes to the debug process can cause toil and frustration for customers and TSE alike.

Jira issue: CRDB-31776