If the rows are not ordered by pos then you need to keep a total for each bin. Consider:

#!/usr/bin/perl
use warnings;
use strict;
use 5.010;

my $binSize = 20;
my %binTotals;

while (<DATA>){
    chomp;
    my ($chr, $pos, $coverage) = split /\t/;

    $binTotals{int(($pos - 1) / $binSize)} += $coverage;
}

printf "%4d %d\n", 20 * (1 + $_), $binTotals{$_} for sort {$a <=> $b} 
+keys %binTotals;

__DATA__
chr    1    2
chr    4    2
chr    7    5
chr    22    5
chr    24    6
chr    38    10
chr    44    10
chr    50    20
chr    57    25
chr    60    30
chr    65    30
[download]

Prints:

  20 9
  40 21
  60 85
  80 30
[download]

and actually that is a lot cleaner than code that assumes the rows ore ordered so doesn't need to accumulate totals for bins:

my $binSize = 20;
my $bin = 0;
my $total = 0;

while ((my $line = <DATA> // '') || defined $bin){
    chomp $line;
    my ($chr, $pos, $coverage) = split /\t/, $line;
    my $thisBin;

    $thisBin = int(($pos - 1) / $binSize) if defined $pos;

    if (! defined $thisBin || $bin != $thisBin) {
        printf "%4d %d\n", 20 * (1 + $bin), $total;
        last if ! defined $thisBin;

        $bin = $thisBin;
        $total = 0;
    }
    
    $total += $coverage;
}
[download]