Non-user-facing work isn't vital. Market velocity is vital. You need to demonstrate you can increase market velocity by investing in non-user-facing work. If your investment in that work slows market velocity then your investment return is negative.
Are you tracking engineering velocity? If so petition to do some internal investment and see if velocity improves afterwards.
Also you should refactor as you go. That isn't even something you break out into a separate task. Clean up messes as you work. Don't leave messes if you can. Code review always and often, and on small changes.
Note: Market velocity is slightly different than user-facing changes. If you have reliability or security failures you are reacting to, fixing them is user facing but not improving your market velocity.
Are you tracking engineering velocity? If so petition to do some internal investment and see if velocity improves afterwards.
Also you should refactor as you go. That isn't even something you break out into a separate task. Clean up messes as you work. Don't leave messes if you can. Code review always and often, and on small changes.
Note: Market velocity is slightly different than user-facing changes. If you have reliability or security failures you are reacting to, fixing them is user facing but not improving your market velocity.