ARROW-17643: [R] Latest duckdb release is causing test failure

[paleolimbot]

#14065 experimented with ways to solve this from Arrow's end; however, the fix in duckdb/duckdb#4712 is probably more robust. Even if that fix doesn't make it in to the next DuckDB release, when that happens this test will start failing again in case we/I forget to check.

[github-actions]

https://issues.apache.org/jira/browse/ARROW-17643

[github-actions]

⚠️ Ticket has not been started in JIRA, please click 'Start Progress'.

[ursabot]

Benchmark runs are scheduled for baseline = b48d228 and contender = 6926672. 6926672 is a master commit associated with this PR. Results will be available as each benchmark for each run completes.
Conbench compare runs links:
[Finished ⬇️0.0% ⬆️0.0%] ec2-t3-xlarge-us-east-2
[Finished ⬇️0.27% ⬆️0.03%] test-mac-arm
[Failed ⬇️0.28% ⬆️0.0%] ursa-i9-9960x
[Finished ⬇️0.14% ⬆️0.0%] ursa-thinkcentre-m75q
Buildkite builds:
[Finished] 69266721 ec2-t3-xlarge-us-east-2
[Finished] 69266721 test-mac-arm
[Failed] 69266721 ursa-i9-9960x
[Finished] 69266721 ursa-thinkcentre-m75q
[Finished] b48d2287 ec2-t3-xlarge-us-east-2
[Finished] b48d2287 test-mac-arm
[Failed] b48d2287 ursa-i9-9960x
[Finished] b48d2287 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

[ursabot]

['Python', 'R'] benchmarks have high level of regressions.
test-mac-arm
ursa-i9-9960x

[pitrou]

Is this actually fixed upstream? I see a build using DuckDB 0.5.1 here and it has the error:
https://github.com/apache/arrow/actions/runs/3095675589/jobs/5010340520#step:7:31839

[paleolimbot]

It looks like no! t I made ARROW-17809 and will look into it today and tomorrow.