Update security-severity scores

[calumgrant]

Run using latest changes to the CWE tooling (https://github.com/github/cwe-scores).

The change is that the score is based on the CVE "Base Score" instead of its "Impact score".

The result of this is that the scores map better to the UI concept of (Low/Medium/High/Critical).

[calumgrant]

The distribution of the scores are as follows:

Language	Low	Medium	High	Critical
All	9	144	332	138
C#	1	22	64	24
C++	2	13	83	33
Golang	0	13	22	10
Java	0	60	95	46
JavaScript	6	37	56	26
Python	0	12	34	9

(Golang is in a separate repo and is not included in this PR).

[aschackmull]

From a quick look through, this looks like an improvement.

[MathiasVP]

These scores reflect my expectations much better 👍! Like @rdmarsh2, I'm curious about why some of the scores are slightly lowered by this change, though.
(It's not a big deal since none of the drops cause a query to change to a lower category.)

[RasmusWL]

👍 For python

[sj]

Thanks @calumgrant! Is it possible to easily generate an overview like this for us to do an extra sanity check:

security-severity \ severity	recommendation	warning	error
low	1	2	3
medium	4	5	6
high	7	8	9
critical	10	11	12

This will allow us to check how many things that we previously gave a high severity (i.e., error) actually end up having a low security-severity.

(cc @AlonaHlobina)