Option 3 should be updated to specify to manually search in the field list to then add additional fields that have the same name but are different actual customfield_ids

this work-around itself is the same work-around you use in Jira Cloud for Sheets which itself suffers from the same bug. Jira Cloud for Sheets is itself not a work-around, the work-around is the same per the steps below in Daria's comment which are for Excel, but the exact same steps are what you use in Jira Cloud for Sheets to work-around this. (You can search for the different fields by the same name if you go to the Fields tab and Edit fields)

Unfortunately this isn't necessarily a minor bug though as this feature is itself how others have identified to find the customfield_id itself in the first place (and, in fact, at the time of writing this comment, that is the accepted answer even though it suffers from this bug and the i.d. you're seeing is wrong)

Have to agree with erika, it is annoying I had one project with Custom Fields export correctly and others not - along with other issues such as the 120 attachments then they all become non readable, not being able to clone a Project to use as a test environment like older versions allowed, for the cloning issue I've had app recommended to use, attempted to used them and they either don't work with Team Managed projects, the ones that say they do, run into errors down track and suggest that you change to Company Managed Projects to get past the issue. Whilst these issues may seem small to Atlassian added together make the application cumbersome to use. to get the desired results in a timely manner. Which overall makes it less attractive to utilizse.

Hi,

Can I please ask for an update, please? This is quite a burning update for us. We can populate data from custom fields through CSV format, but for each project, the same custom fields gets a different column and you need to merge the data manual to be able to provide solid reporting. Additionally CSV format does not recognize signs from other languages hence converts them into different signs, which again speaks against the CSV export. 

Hello everyone,
I’ve tried to perform the same steps as in issue.

There are 2 Next-gen Jira Projects (TESTD2, TESTD3) in my Test Jira, each has ‘Label Test’ Custom Field (drop-down):

For each Project Issue with filled ‘Label Test’ Field is created. 

When open separate Issue, each has only one ‘Label Test’ Field. 
When search for those issues in Jira, each has two values for ‘Label Test’ Field (for each Project), for example:

‘TESTD2 US 1’ has ‘labelTESTD2_1’ value for TESTD2 Project and ‘None’ for TESTD3 Project;
‘TESTD3 US 1’ vice versa has ‘None’ value for TESTD2 Project and ‘labelTESTD3_1’ value for TESTD3 Project.
TESTD2 goes 1st in value list for ‘Label Test’ Field.

When export in Excel, only 1st value is shown for the column, in my case value for TESTD2 Project. That is why issues for TESTD3 Project have empty ‘Label Test’ Field:

Also when select Columns in Jira, this Field is mentioned just once in the list (no indication to which Project that field belongs):

Also to have Custom Field values for all Issues, User may select both these Fields in Add-In and get the following result:

Workaround: in Excel File created using 'open in Microsoft Excel' - select all Fields with the same names in Add-In and get data again. Or give different names to Custom Fields in different Projects (as mentioned in current Ticket).

Internal Ticket is created and will be taken in work basing on Team workload.