Partial 325 Fix P2 - Updated vectvshash.png Added

[GalinaP7]

Description

Follow-up to #1137 - adds the replacement graph for fig-vectvshash-cpp, which was commented out in that PR since the old
image reflected the previous (incorrect) cumulative-timing measurement.

fixes #325

What's fixed

New chart generated from corrected code output

• Ran the corrected C++ example (std::chrono-based, per-operation average timing) locally across the full range (10,000 to 990,000 elements, step 20,000).
• Built a new vectvshash.png in Excel from that real output, replacing the old image which was based on cumulative search time (0–50 second scale) rather than per-operation average time (microsecond/nanosecond scale).
• Converted vector search time to microseconds for readability (raw values were in the 1e-07 to 1e-05 second range, which rendered as hard-to-read scientific notation on a linear axis).
• Axis titles, legend labels (Vector / Hash Table), and marker shapes (circle for vector, triangle for hash table) were set to match the style of the original figure.

Figure re-enabled

• Uncommented the <figure xml:id="fig-vectvshash-cpp"> block in HashTableAnalysis.ptx.
• Restored the <xref ref="fig-vectvshash-cpp"/> in the results paragraph that references it.
• Updated the figure's <description> text to describe the actual new image: vector climbs to approximately 69 microseconds at 1,000,000 elements, hash table stays flat at approximately 2–3 nanoseconds throughout (previously said "45 seconds," which
  reflected the old, incorrect measurement).

Tested on local build

Reviewed by @harrisonj2-v

[wrigjl]

Am I missing something (quite possible)? I see that you've updated the text, but I don't see the new image in the PR (missing a "git add"?).

[GalinaP7]

That's completely my fault! I forgot to attach the PNG.

[wrigjl]

Heh, happens to everyone =)

[pearcej]

This is a big improvement! However, I have a minor nit. I would rather the xml:id be fig-vectvshash2-cpp rather than fig-vectvshash-cpp2 because I think semantically the change is not to the language, but the figure.

[pearcej]

@wrigjl Thoughts on this and part 1 (#1137)?