rm TODO comment about integrating detectLateRejectionHandling with clearStack

In theory, resetting clearStack's inline call count every time late
rejection handling does a setTimeout should reduce the performance penalty
in some environments.  In practice, it doesn't:

* In Chrome and FF, late rejection handling has no measurable penalty.
* In Safari, resetting the inline call count actually slows things down
  considerably(!).
* In Node, clearStack doesn't use setTimeout so there is no benefit.

lib/jasmine-core/jasmine.js

@@ -11614,7 +11614,6 @@ getJasmineRequireObj().TreeRunner = function(j$) {
       return {
         fn: done => {
           // setTimeout is necessary to trigger rejectionhandled events
-          // TODO: let clearStack know about this so it doesn't do redundant setTimeouts
           this.#setTimeout(function() {
             globalErrors.reportUnhandledRejections();
             done();

src/core/TreeRunner.js

@@ -192,7 +192,6 @@ getJasmineRequireObj().TreeRunner = function(j$) {
       return {
         fn: done => {
           // setTimeout is necessary to trigger rejectionhandled events
-          // TODO: let clearStack know about this so it doesn't do redundant setTimeouts
           this.#setTimeout(function() {
             globalErrors.reportUnhandledRejections();
             done();