The performance improvement with bulk loading is an order of magnitude faster. The
ehcacheperf (Spring Pet Clinic) project now has a bulk-load test that shows the performance improvement for using a Terracotta cluster. Consider also that multi-threading is likely to improve performance.
Why not run in bulk load mode all the time? - Terracotta clustering provides consistency, scaling and durability. Some applications might require consistency. However, for reference data it might be acceptable to run a cache permanently in inconsistent mode.