Before this commit, we used FirAccessorSymbol to emulate synthetic properties.
These symbols were generated in Java use-site scope.
Now, we use synthetic scope instead which is above MemberScopeTowerLevel.
This is more performance-friendly and does not require override matching.
However, accessor symbols should be used in situation when Java accessor
overrides Kotlin property which is broken in this commit
(that's why MapEntry test is corrupted here).
Also, we should not create synthetics for pure Kotlin accessors
(that's why javaAccessorConversion test is corrupted here).
Really, this commit implements early J2K mapping for all Java types.
It's questionable and probably wrong at least for super-types,
because, for example, we cannot resolve spliterator() in classes
derived from java.lang.Iterable
Mikhail Glukhikh