Revert change to 1D wide data handling in categorical scatterplots

[mwaskom]

This reverts a change introduced at the beginning of the categorical plot modernization process (#2429), where 1D wide data (internally called "flat" data, although I am starting to prefer "thin") was pivoted like in the relational plots, with the series index going to x and values going to y.

While this makes sense for functions that aggregate (barplot/pointplot), I don't think it make sense for the distributional functions, so this replaces the original behavior:

sns.swarmplot(tips["total_bill"], orient="h")

It adds the slight improvement that the categorical axis would previously have a "0" tick label, whereas now the tick is unlabeled, akin to when you pass the data directly to x or y.

[codecov]

Codecov Report

Merging #2829 (a01af81) into master (dce3150) will increase coverage by 0.00%.
The diff coverage is 100.00%.

@@           Coverage Diff           @@
##           master    #2829   +/-   ##
=======================================
  Coverage   98.05%   98.05%           
=======================================
  Files          68       68           
  Lines       22065    22066    +1     
=======================================
+ Hits        21636    21637    +1     
  Misses        429      429           

Impacted Files	Coverage Δ
seaborn/categorical.py	98.33% <100.00%> (+<0.01%)	⬆️
seaborn/tests/test_categorical.py	99.77% <100.00%> (-0.01%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update dce3150...a01af81. Read the comment docs.