Email (record): On the eigenvalue effective size of structured populations