Redefine default allocations following alex suggestions

Change defaults to:

• gracedb.resources.cpu: 4 (or 2)
• gracedb.resources.memory: 8Gi (or 4Gi)
• gracedb.unicorn.worker: 9 (or 5)

From Alex:

• gracedb's development VM's are all dual-core with 4GB ram.
• the dev and test ec2 nodes in aws are similarly dual-core with 4GB ram (granted there's a cluster of three of them)
• the production nodes are 4-core with 8gb ram, and aws is constantly complaining about how they're over-provisioned

Actual default allocations for GraceDB are:

gracedb:
  image: "containers.ligo.org/computing/gracedb/server:gracedb-2.31.1"
  storage:
    capacity: 10Gi
  resources:
    cpu: 6
    memory: 8Gi

  gunicorn:
    worker: 13
    threads: 1
# Number of gunicorn workers
  # 2*CPU + 1 (recommendation from Gunicorn documentation)
  # NOTE: it was found in extensive testing that threads > 1 are prone
  # GUNICORN_WORKERS = {{ .Values.gracedb.gunicorn.worker }}
  # GUNICORN_THREADS = {{ .Values.gracedb.gunicorn.threads | default 1 }}