Fix kernel and launch dimensions for coo2dense function in opencl, cuda, and oneapi backends.

Description

• Is this a new feature or a bug fix?: Bug fix
• Why these changes are necessary?: Fixes incorrect behavior of af::dense
• Potential impact on specific hardware, software or backends: affects OpenCL, CUDA, oneAPI
• New functions and their functionality.: None
• Can this PR be backported to older versions?: Only changes in OpenCL and CUDA backends can be backported
• Future changes not implemented in this PR.: None

Fixes: #3542

Changes to Users

• No additional changes added to build
• No action required by the user

Checklist

• Rebased on latest master
• Code compiles
• Tests pass
• Functions added to unified API
• Functions documented