Tests regression after the proj4j 1.1.4 upgrade

[pomadchin]

Describe the bug

github-test-1  | [info] OverzoomingValueReaderSpec:
github-test-1  | [info] OverzoomingValueReader
github-test-1  | [info] - should work for simple case *** FAILED *** (101 milliseconds)
github-test-1  | [info]   java.lang.NullPointerException:
github-test-1  | [info]   at org.locationtech.proj4j.datum.Grid.equals(Grid.java:393)
github-test-1  | [info]   at java.util.AbstractList.equals(AbstractList.java:523)
github-test-1  | [info]   at org.locationtech.proj4j.datum.Datum.isEqual(Datum.java:205)
github-test-1  | [info]   at org.locationtech.proj4j.CoordinateReferenceSystem.equals(CoordinateReferenceSystem.java:134)
github-test-1  | [info]   at geotrellis.proj4.CRS.equals(CRS.scala:161)
github-test-1  | [info]   at geotrellis.proj4.CRS.equals$(CRS.scala:159)
github-test-1  | [info]   at geotrellis.proj4.WebMercator$.equals(WebMercator.scala:21)
github-test-1  | [info]   at geotrellis.layer.package$CRSWorldExtent$.worldExtent$extension(package.scala:61)
github-test-1  | [info]   at geotrellis.layer.ZoomedLayoutScheme.levelForZoom(ZoomedLayoutScheme.scala:104)
github-test-1  | [info]   at geotrellis.store.OverzoomingValueReader$$anon$1.<init>(OverzoomingValueReader.scala:39)
github-test-1  | [info]   ...

This could be the proj4j regression or the GT misbehavior. It was revealed here: #3452

[pomadchin]

Closed via locationtech/proj4j#78 and is used here #3452