Minor wording changes in design doc

sycl/doc/design/CommandGraph.md

@@ -438,23 +438,23 @@ Level Zero:
 Future work will include exploring L0 API extensions to improve the mapping of
 UR command-buffer to L0 command-list.
 
-#### Copy engine
+#### Copy Engine
 
-For performance considerations, Unified-Runtime uses different Level-zero
-command-queues to submit compute kernels and memory operations when the device
-has a dedicated copy engine. To take advantage of the copy engine when
-available, the Graph workload can also be split between memory operations and
-compute kernels. To achieve this, two Graph workload command-lists live
-simultaneously in a command-buffer.
+For performance considerations, the Unified Runtime Level Zero adapter uses
+different Level Zero command-queues to submit compute kernels and memory
+operations when the device has a dedicated copy engine. To take advantage of the
+copy engine when available, the graph workload can also be split between memory
+operations and compute kernels. To achieve this, two graph workload
+command-lists live simultaneously in a command-buffer.
 
-When the command-buffer is finalized, memory operations (e.g. Buffer copy,
+When the command-buffer is finalized, memory operations (e.g. buffer copy,
 buffer fill, ...) are enqueued in the *copy* command-list while the other
-commands are enqueued in the general command-list. On submission, if not empty,
-the *copy* command-list is sent to the main copy command-queue while the general
+commands are enqueued in the compute command-list. On submission, if not empty,
+the *copy* command-list is sent to the main copy command-queue while the compute
 command-list is sent to the compute command-queue.
 
 Both are executed concurrently. Synchronization between the command-lists is
-handled by Level-Zero events.
+handled by Level Zero events.
 
 ### CUDA