chore: require value in setAttribute

[mwear]

Which problem is this PR solving?

• Fixes setAttribute value should be required #1848

Short description of the changes

• This PR makes the value required in Span#setAttribute and updates several call sites where potentially undefined values were being passed in.

[codecov]

Codecov Report

Merging #1851 (98637bd) into main (f4e7115) will increase coverage by 0.37%.
The diff coverage is 100.00%.

@@            Coverage Diff             @@
##             main    #1851      +/-   ##
==========================================
+ Coverage   92.35%   92.72%   +0.37%     
==========================================
  Files         157      174      +17     
  Lines        5104     6038     +934     
  Branches     1085     1283     +198     
==========================================
+ Hits         4714     5599     +885     
- Misses        390      439      +49     

Impacted Files	Coverage Δ
...s/opentelemetry-instrumentation-fetch/src/fetch.ts	99.23% <100.00%> (ø)
...emetry-instrumentation-xml-http-request/src/xhr.ts	96.82% <100.00%> (ø)
packages/opentelemetry-web/src/utils.ts	94.66% <100.00%> (ø)
...ry-exporter-collector/src/CollectorExporterBase.ts	92.45% <0.00%> (ø)
...kages/opentelemetry-web/src/StackContextManager.ts	97.05% <0.00%> (ø)
...ry-context-zone-peer-dep/src/ZoneContextManager.ts	85.18% <0.00%> (ø)
...ages/opentelemetry-exporter-collector/src/types.ts	100.00% <0.00%> (ø)
...lemetry-exporter-collector/src/transformMetrics.ts	95.00% <0.00%> (ø)
... and 10 more

[dyladan]

I agree with @Flarna that != will catch more cases than !== without overly complicating the code, but this is a matter of preference. Since null/undefined are technically undefined behavior it might not matter that much.

[MSNev]

nit: !== undefined

[dyladan]

I think we may end up stuck in a loop #1851 (comment)