Fix bug where loading spoor-id from the cookie would silently fail

[JakeChampion]

spoor-id is stored directly as a string and not an json object, which meant json.parse would throw an error and we would not be migrating the old spoor-id value to localstorage and instead would mint a new spoor-id, which is not what we want to do.

[origamiserviceuser]

o-tracking bundle size difference from 3.0.3 to 5117614f778630b4bc377e17b1921b7799eb0a4e No bundle size differences found.

[Lily2point0]

I clicked 'approve' by force of habit, I actually had a comment first but I don't know how to remove my approval 😅

[notlee]

Flying brain dump before leaving for lunch: If the cookie is not removed won't any data set to local storage be ignored on future visits? Maybe the content of the cookie should be emptied except ids and the check be based on that?

[JakeChampion]

Flying brain dump before leaving for lunch: If the cookie is not removed won't any data set to local storage be ignored on future visits? Maybe the content of the cookie should be emptied except ids and the check be based on that?

Good idea, I'll implement that 👍

Just for refence, I check o-tracking v2 and there are 3 stores being used:

• A store for the queue, which will use localStorage by default and fallback to cookies if localStorage does not exist
• A store for the session, which will use localStorage by default and fallback to cookies if localStorage does not exist
• A store for the user, which always uses cookies and never localStorage

[origamiserviceuser]

o-tracking bundle size difference from 3.0.3 to a5b74598a330c49b0f3f3f0f246b95c1b6ab1a59 No bundle size differences found.

[origamiserviceuser]

o-tracking bundle size difference from 3.0.3 to edf82b0a93e4c6ccd21ebfd193cb781e44534c13 No bundle size differences found.

[origamiserviceuser]

:tada: This PR is included in version v3.0.4 :tada: