Fix file corruption and straighten out vented/unvented at TMP

[bpbond]

OK, whew, this took some investigative work.

We had several problems:

• AquaTROLL data being incorrectly flagged (and then dropped in L2) because the bounds were set based on vented data
• Many edited raw files (due to previous fixes such as TMP AQ600 density data wrong columns issue #245 Non-vented and vented troll issue - GCW only #342 -- not sure where the problem originated) were not being read correctly -- read_datalogger_file() in L0 was generating a warning that was being ignored, because there were more columns than column names

This PR:

• Reverts the back-corrections made to the AquaTROLL data at TMP-- fyi @wilsonsj100
• Fully implements the gw-pressure-vented and gw-pressure-unvented research names merged in Split gw-pressure into vented and unvented #444
• Adds design table mapping for the trolls that changes over from unvented to vented in September 2024
• Redoes the column-swap fix originally made in TMP AQ600 density data wrong columns issue #245
• Addresses the OOB issue noted by @stephpenn1 in Unvented Aquatroll pressure incorrectly flagged as OOB in L1 #443

The result is I think what we want:

Closes #446

[bpbond]

One disadvantage is that people may go back to asking @wilsonsj100 about why the troll pressure data readings are so high. But hopefully the gw-pressure-vented and gw-pressure-unvented variable names, and their accompanying documentation, will mitigate this.

[bpbond]

This may also fix #445 -- it would make sense. But I have not confirmed that.

[stephpenn1]

This looks great, thanks for your detailed work here. Fingers crossed it fixes #443