ARROW-15822: [C++] Cast duration to string (thus CSV writing) not supported

[benibus]

See: ARROW-15822.

[github-actions[bot]]

https://issues.apache.org/jira/browse/ARROW-15822

[benibus]

Might be good to go. I can also confirm that the CSV tests fail without these changes.

[ursabot]

Benchmark runs are scheduled for baseline = a1b161e85a4dce497d65325db945f72bbd20b729 and contender = b41bd80187d88f5187a3dc7c42444fbbcca6f7a8. b41bd80187d88f5187a3dc7c42444fbbcca6f7a8 is a master commit associated with this PR. Results will be available as each benchmark for each run completes. Conbench compare runs links: [Finished :arrow_down:0.0% :arrow_up:0.0%] ec2-t3-xlarge-us-east-2 [Failed :arrow_down:0.0% :arrow_up:0.0%] test-mac-arm [Finished :arrow_down:0.82% :arrow_up:0.0%] ursa-i9-9960x [Finished :arrow_down:1.75% :arrow_up:0.29%] ursa-thinkcentre-m75q Buildkite builds: [Finished] b41bd801 ec2-t3-xlarge-us-east-2 [Failed] b41bd801 test-mac-arm [Finished] b41bd801 ursa-i9-9960x [Finished] b41bd801 ursa-thinkcentre-m75q [Finished] a1b161e8 ec2-t3-xlarge-us-east-2 [Failed] a1b161e8 test-mac-arm [Finished] a1b161e8 ursa-i9-9960x [Finished] a1b161e8 ursa-thinkcentre-m75q Supported benchmarks: ec2-t3-xlarge-us-east-2: Supported benchmark langs: Python, R. Runs only benchmarks with cloud = True test-mac-arm: Supported benchmark langs: C++, Python, R ursa-i9-9960x: Supported benchmark langs: Python, R, JavaScript ursa-thinkcentre-m75q: Supported benchmark langs: C++, Java