fix: add ability to import RecordTypeId when RecordType.Name is present

[WillieRuemmele]

What does this PR do?

if you export with a similar command data tree export --query "SELECT RecordType.Name, Name FROM Account
data:tree:import can now resolve the RecordType sub-reference correctly

NUTs weren't possible due to a manual step in the target org 😦

What issues does this PR fix or reference?

@W-4472514@

[shetzel]

QA
org1:

• create RecordType on Account object
• create a few Account records
• sf data tree export --plan --query "SELECT Recordtype.Name, Name FROM Account"
• sf project retrieve start -m CustomObject:Account

org2:

• sf project deploy start -m CustomObject:Account
• In org Setup UI, make the RecordType created in org1 available to Admin user
• bin/run.js data tree import --plan Account-plan.json - ensure success
• Also ran sf data query -q "select Recordtype.Name, Name from Account" and saw the imported data

[thvd]

@WillieRuemmele Trying to understand here, don't get me wrong, very happy with this new improvement.

But what I don't understand is that this is implemented were we use the RecordType.Name field, what i would have found more logical is to use a combination of Namespace and DeveloperName field.
Will this current solution still work were record type are duplicated?