Storing non-native values such as dates and timestamps in a VARIANT column in Snowflake can lead to slower query performance and increased storage consumption. This is because these values, when stored as VARIANT, are often converted to string format. As a result, operations on these values may not be as efficient as they would be if the data were stored in a native date or timestamp data type, leading to slower performance. Additionally, storing these non-native values as strings can require more storage space compared to their native formats.