add error message grouping at the end of fetch step

[TheBlindHawk]

Describe the user story

I was upgrading multiple libraries from a repository, and after checking that they were working I deployed them for staging.
The staging branch failed to deploy due to a yarn error (YN0041) which took a lot more time to find that it could have otherwise.
The reason it took so long is because the error message was hidden deep within multiple other "non-error" messages like this:

➤ YN0000: ┌ Fetch step
➤ YN0013: │ @antfu/ni@npm:0.21.5 can't be found in the cache and will be fetched from the remote registry
➤ YN0013: │ @aws-crypto/crc32@npm:3.0.0 can't be found in the cache and will be fetched from the remote registry

    // a few hundred/thousand of rows with YN0013

➤ YN0041: │ ... Invalid authentication (as an unknown user)

    // a few hundred/thousand of rows with YN0013

➤ YN0013: │ zip-stream@npm:4.1.0 can't be found in the cache and will be fetched from the remote registry
➤ YN0013: │ zod@npm:3.21.4 can't be found in the cache and will be fetched from the remote registry
➤ YN0000: └ Completed in 37s 463ms
➤ YN0000: Failed with errors in 37s 696ms

At first I thought that the build had failed without outputting any errors, but later came to the conclusion that that could not be the case and found the error.

Describe the solution you'd like

Would it be possible to output error that caused the install to fail right before/after the Failed with errors line?

➤ YN0000: Failed with errors in 37s 696ms
➤ YN0000: ┌ Cause of failure
➤ YN0041: │ ... Invalid authentication (as an unknown user)
➤ YN0000: └ Failed with errors in 37s 696ms

maybe it could be an optional yarn install --relog-errors.

Describe the drawbacks of your solution

It is probably not a very required feature, with only a few use case scenarios I can think of, as in most cases the error log is quite easy to find.
In case it is not set as a command option it would cause the error message to be printed twice, which could be a bit confusing in the absence of other logs.

➤ YN0000: ┌ Fetch step
➤ YN0041: │ ... Invalid authentication (as an unknown user)
➤ YN0000: └ Completed in 37s 463ms
➤ YN0000: Failed with errors in 37s 696ms
➤ YN0000: ┌ Cause of failure
➤ YN0041: │ ... Invalid authentication (as an unknown user)
➤ YN0000: └ Failed with errors in 37s 696ms

Describe alternatives you've considered

I think other alternatives would be to stop the logs when the error occurs or to have the error logged at the end , instead of when it occurs:

➤ YN0000: ┌ Fetch step
➤ YN0013: │ @antfu/ni@npm:0.21.5 can't be found in the cache and will be fetched from the remote registry
➤ YN0013: │ @aws-crypto/crc32@npm:3.0.0 can't be found in the cache and will be fetched from the remote registry

    // a few hundred/thousand of rows with YN0013

➤ YN0041: │ ... Invalid authentication (as an unknown user) 
    // stop logging anything after here, or log this line at  the end of every other YN0013 log
➤ YN0000: └ Completed in 37s 463ms
➤ YN0000: Failed with errors in 37s 696ms