Do Methodological Choices in Environmental Modeling Bias Rebound Effects? A Case Study on Electric Cars