We have NetApp storage as well and our core is SQL Server and we've made a similar decision. Our main issue was with the database quiescing that any snap product introduces to the SQL environment (this is a Microsoft VSS issue in the end, NOT a NetApp issue).
In these cases, it quiesces all of the databases, serially, before taking a backup. This causes our databases to "hold" writes for several seconds every time we take a backup. In our testing, the "holding" starts when it quiesces the first database on the SQL instance, then continues to hold writes until the final database on that instances is quiesced. We have over 30 databases on some servers, and it takes 1 - 7 seconds to queisce a database. This may not pose a problem if your database servers house more "warehouse" like data, that is only queried periodically. However, as in our case, we're more OLTP, where our databases are serving up 1000s of transactions per second 24x7.
We decided to house and maintain the SQL Server native Full/Differential/Log backups. We do NOT snapshot them at this time though for the same reasons mentioned above. Every night we do a full, with a diff every few hours, and frequent transaction log backups. This amounts to huge amounts of change, which makes snapping those volumes unmanageable.