[BUG] Double bonds rendered out of plane (CID: 243, 2244)

[dxdc]

Describe the bug
In some cases, the double bond for a carbonyl is being drawn out of plane. I'm wondering if -- inadvertently -- this is related to the getGoodCross function we've been discussing in #455.

To Reproduce
View the carbonyl rendered with the CID examples below (243, 2244). In the case of 2244, one of the carbonyls is correct and the other is not. I'm not sure if there are other examples, but these are at least two.

Screenshots

CID: 243 (Benzoic Acid)

CID: 2244 (Aspirin)

[dkoes]

Yup, an updated getGoodCross would change this behavior.

[dkoes]

Sorry for the radio silence - been swamped. The above push should standardize these bonds (was actually unrelated to getGoodCross).
https://3dmol.org/tests/auto/generate_test.cgi?test=testdoubles