I beg to differ with you Colin.  However remember that it is a max of 3 per disk in a diskset (per MS.)  I would also advise the use of 'Current Disk Queue' not average.  Obviously if your Avg is through the roof then you have identified an issue but Current will give you the point in time view of what is going on.  You can correlate a lot of things that way.  If your database drive Current Disk queue is 125 on a R5 (4 disks) you are at 10 times the acceptable level.  If during this time all of your queries are running at 0 ms then it is not appearing to be a related issue.  If however your queries are all running several minutes or more then this could correlate.  Another thing to keep in mind is that there are more factors than disk queue and query performance.  Is all of your work happening at one time that could be spread out?  Is there blocking? Does the machine have enough memory for the current used dataset?

Performance monitoring can be lengthy and provide you with many headaches, however it can also provide you with the data needed to rationalize a large investment or to prove to your Network Ops team that the bottleneck in the users percieved performance problems is not on your database server but somewhere beyond it.