merlyn is right. Rose is wicked fast. It's fast enough that you're unlikely to notice too much of a difference and it's easy to fine-tune, if needed. The wins with Rose (and its clean design), easily outweigh the losses.

You can see the benchmarks for yourself.

When you work with many tables and gigabytes of data, you have to fine-tune many queries. The last time I had these problems, this tuning was not that easy because an ORM is kind of a blackbox - RDO probably less than others. I ended up with ugly code that was not more maintainable than plain DBI code. Well, probably I am someone who does silly things with ORMs ;)

But I really like RDO, so I'd interested in reading more about things you should consider (probably starting with the db schema) when you go beyond little webapp databases ;)

You're correct: complicated queries on large databases are no place for an ORM. Rose is pretty fast for simple-to-medium queries, but complex ones always require hand-tweaked SQL. You can run hand-written SQL through Rose though (or others), and it can still provide a convenient place to organize your custom queries and deal with your database handle/statement caching.

I would also add that ORMs are never good for bulk updates. You can totally crush the speed of an ORM for a bulk insert/update by writing some simple DBI code or by using your database's bulk loading features, e.g. multi-row insert in MySQL.