I think you are missing a hyphen in --skip-unmapped, but --echo-map-id-uniq is working correctly.

The ID field includes the exon number, in this case, so an ID like TDRD10:10 is treated (correctly) as a distinct string from TDRD10:2.

If you want a tab between range and IDs, use --delim '\t' with bedmap. This replaces the pipe character ('|') with a tab.

Otherwise, you could collapse the IDs by prefix by post-processing the current output with Perl or Python, etc.

For instance:

#!/usr/bin/env perl

use strict;
use warnings;

while (<STDIN>) {
    chomp;
    my ($rangeStr, $idsStr) = split("|", $_);
    my @ids = split(";", $idsStr);
    my $prefixHash;
    foreach my $id (@ids) {
        my ($prefix, $exon) = split(":", $id);
        if (! defined $prefixHash->{$prefix}) { $prefixHash->{$prefix} = (); }
        push $exon, @($prefixHash->{$prefix});
    }
    print STDOUT $rangeStr."\t";
    foreach my $prefix (sort keys %{$prefixHash}) {
        print STDOUT "$prefix:".join(",", sort @{$prefixHash->{$prefix}}).";";
    }
    print STDOUT "\n";
}

Untested, but this should perhaps spit out something more readable, like:

chr1    154542723    154542853    TDRD10:3,8;
chr1    154543654    154544647    TDRD10:2,3,4,5,9,10;

Note that you would want to use --skip-unmapped to prepare input for this script. Or else you'll likely get one or more lines with an empty fourth field where there are no mapped elements.