feat: remove JS benchmarks and reduce depth for faster binary_trees/fib runs

Switch binary_trees max depth from 14 to 12 across all languages and reduce fib argument from 30 to 28, dropping JS entirely since its performance is incomparable. Update expected output patterns in run_bench script to match new tab-free format and use best-time comparison instead of mean.

benchmark/fib.wren

@@ -9,7 +9,7 @@ var fib = fn(n) {
 var start = OS.clock
 var i = 0
 while (i < 5) {
-  io.write(fib.call(30))
+  io.write(fib.call(28))
   i = i + 1
 }
 io.write("elapsed: " + (OS.clock - start).toString)