I repeated the OP analysis with 3-month holds, and got similar results. For mega-caps, the top market cap stock has lower returns, and adding a earnings growth sort increases returns. The mega-cap with the highest EG has lower returns.

                           19851231  19570301  from
final sort      N     OP   19981231  20220126   to
  MktCap      Top 1   7%      10         7
  MktCap      Top 3   16%     16        11
  MktCap      Top 5   15%     17        11
  MktCap    6 to 10   19%     20         9
  MktCap    11 to 30  20%     21        10
 
                           19880302  19880302  from
final sort      N     OP   19981231  20220126   to
    EG        Top 1   13%      4         5
    EG        Top 3   30%     21        11
    EG        Top 5   26%     24        13     
    EG      6 to 10   21%     21        14
    EG      11 to 30  18%     20        11
 
    EG         #2     31%     22         8
    EG       2 to 3   37%     29        13
    EG       2 to 4   32%     30        14
    EG       2 to 6   28%     29        14     

https://gtr1.net/2013/?~SP100EG:h63::sp500.a:et1:rank%28clas...

I did additional backtests with each holding 5 stocks. {SP100} highest CAGR was positions 2 to 6. {SP100EG} highest CAGR was positions 4 to 8.

{SP100} highest CAGR is param0=6, using a final sort of:
MktCap top param0
MktCap bottom 5

param0  CAGR  GSD
  5     10.5  17.7
  6     11.0  18.0
  7     9.4   18.4
  8     9.1   18.7
  9     9.4   19.0
  10    8.9   19.2
  15    9.4   19.5
  20    8.9   19.1
  30    10.0  19.4
  50    10.2  21.0
  100   9.3   21.2

{SP100EG} highest CAGR is param0=8, using a final sort of:
EG top param0
EG bottom 5

param0  CAGR  GSD
  5     13.1  25.2
  6     14.1  25.1
  7     14.0  25.1
  8     14.1  24.9
  9     14.0  24.7
  10    13.9  24.4
  15    8.9   23.2
  20    11.4  21.1
  30    12.1  19.6
  50    9.7   18.1
  100   9.9   24.2