TST: Fixed issues that need tests

[phofl]

• Series.rmod called with scalar returns inconsistent results based on size of series pandas-dev/pandas#29602
• pd.where OverflowError with large numbers pandas-dev/pandas#31687
• df autosorts rows on construction with OrderedDict of OrderedDicts pandas-dev/pandas#18166
• Different behavior of to_replace method between pandas version 0.23.4 and 0.24.2 by changing dtype of series pandas-dev/pandas#25797
• Float conversion makes fillna lossy pandas-dev/pandas#24537
• read_fwf: skip_blank_lines does nothing pandas-dev/pandas#22693
• DataFrame.groupby()[col].transform() tries to convert result to original column type pandas-dev/pandas#22243
• Some groupby methods keep null grouping keys pandas-dev/pandas#21755
• Unexpected results when filtering with .isin (some fields contain python datastructures) pandas-dev/pandas#20883
• API: Should factorize(categorical) return a Categorical for uniques? pandas-dev/pandas#19721
• BUG: Assignment to a .loc view of a naive datetime column changes its dtype to object pandas-dev/pandas#49837
• column names with leading underscores with df.itertuples cause AttributeError when accessed pandas-dev/pandas#23848 (column names with leading underscores with df.itertuples cause AttributeError when accessed pandas-dev/pandas#23848 (comment))
• BUG: MultiIndex intersection doesn't work with categorical dtypes pandas-dev/pandas#49337
• BUG: ValueError converting dense categorical series to sparse when fill_value not in series pandas-dev/pandas#49987
• Substraction with UInt64 series resulting in negative values gives TypeError pandas-dev/pandas#22023
• Series.replace fails to replace value pandas-dev/pandas#32075
• Unexpected behavior in cut() with nullable Int64 dtype pandas-dev/pandas#30787

[codamuse]

I plan on working through these and closing them out. Opened two draft PR's (pandas-dev/pandas#50060 and pandas-dev/pandas#50061). I've asked in #29602 as well so I apologize for the double but since this will happen with all of the above, should I be trying to get fails on these tests on each of the corresponding versions?