remove superfluous conditional

[waitfreemaxi]

The if statement removed always evaluates to true by the following lemma:

Lemma 1:
The confirmation count of the N-th recent lockout (zero-index) is greater than N.

Proof by contradiction: If the Nth recent lockout has a confirmation count less than or equal to N, then, due to the monotonicity of confirmation counts, the (N-1)th most recent lockout must be less than or equal to (N-1). Applying induction, the first most recent lockout (index = 0) must have a confirmation count less less than or equal to 0. However, it is impossible for a vote to have a confirmation count of 0. So there is a contradiction.

[waitfreemaxi]

Alternatively, it's probably faster and cheaper to just check the lockout threshold every time >.<, instead of iterating through a 31 element tower starting from the vote with the highest confirmation count

[waitfreemaxi]

@AshwinSekar

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 82.7%. Comparing base (b711bda) to head (6284aee).
Report is 13 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff            @@
##           master    #1449     +/-   ##
=========================================
- Coverage    82.7%    82.7%   -0.1%     
=========================================
  Files         871      871             
  Lines      370376   370374      -2     
=========================================
- Hits       306561   306540     -21     
- Misses      63815    63834     +19     

[AshwinSekar]

thanks for the change.
As for your comment about improving the check, it could be possible if we had an index into the threshold_vote slot in the old tower. Remember we want to check threshold only if the threshold vote slot has doubled its lockouts, so we can't just compare the 4 / 8 deep slot to see if it changed.