The computing capacity is the number and size of machines we provide to you. The more and larger you have, the more you can do with it and the more it costs. It corresponds to the raw computing power you have available.
The service scale is free, however. It corresponds to the number of times your service is run concurrently. It's up to you to tell where you want the computing power to be used for.
The following screenshot shows a service with its number of instances set to 10:
If you need to scale your application, you need to understand which service needs more power and scale this service up. If the capacity is at its limit, you will also need to increase the cluster capacity.
The following screenshot shows 2 services, a front app with 2 instances and a worker service with 10 instances: