Yes of course, this is completely normal and expected.
The purpose of the RAID is exactly to prevent disks to be read/written separately, they should be always managed together, in the form of the RAID array. The controller would be really confused if we'd want to perform sector-by-sector read (and especially write) test of a RAID member, this would immediately cause synch issues.
With Hard Disk Sentinel, yes, you can monitor the individual disk drives and can also test them INDIVIDUALLY by the Disk menu -> Short self test, Extended self test functions. These hardware self tests are performed "inside" the hard disks, with no real data transfer, during idle periods, when the disk drive does not perform any other commands. So these tests are possible and does not cause any troubles in the operation of the RAID controller.
Also, exactly to verify the stability, integrity of the complete RAID array, you can perform Disk menu -> Surface test functions on the COMPLETE RAID array. This is a good way to verify and confirm the operation of the complete array and the storage subsystem - as this surface test tests both the disks and the controller plus possible additional hardware like backplanes, enclosures, etc.
So this is absolutely not a limitation, exactly the opposite: this way you can have the freedom to test differently to ensure proper operation.