Racket*:several updates

[Antigen-1]

• Add a Racket_concurrency version, which is generally faster than the Racket_compiled and Racket_parallel version.
• raco demod is used for speeding up the Racket_compiled and Racket_concurrency version. The Racket_parallel version cannot take the advantage of raco demod , for place always dynamically requires the enclosing module. That certainly leads to relatively poorer performance.
• The position of --no-compiled flag for the Racket version is fixed, which accounts for intolerably poor performance due to the absence of compiled code.

[Antigen-1]

> cat benchmark.toml | ./run_benchmark Racket_parallel Racket_compiled ChezScheme Racket_concurrency Python3

Running Python3...Ok(BenchmarkResult { topavg: 257212070, avg: 271356826, min: 255825929, max: 319656816, mdev: 21382204, count: 20 })
Running Racket_compiled...Ok(BenchmarkResult { topavg: 591436396, avg: 600609433, min: 585715922, max: 635955370, mdev: 13392069, count: 20 })
Running Racket_concurrency...Ok(BenchmarkResult { topavg: 568573995, avg: 579523443, min: 558149724, max: 602222904, mdev: 12774747, count: 20 })
Running Racket_parallel...Ok(BenchmarkResult { topavg: 643202393, avg: 663995665, min: 620624700, max: 709191822, mdev: 25864078, count: 20 })
                 Python3: top:  257.21, min:  255.83, avg:  271.36, max:  319.66, mdev:   21.38, cnt:  20
              ChezScheme: top:  541.48, min:  484.82, avg:  664.69, max:  867.93, mdev:  140.80, cnt:  20
      Racket_concurrency: top:  568.57, min:  558.15, avg:  579.52, max:  602.22, mdev:   12.77, cnt:  20
         Racket_compiled: top:  591.44, min:  585.72, avg:  600.61, max:  635.96, mdev:   13.39, cnt:  20
         Racket_parallel: top:  643.20, min:  620.62, avg:  664.00, max:  709.19, mdev:   25.86, cnt:  20