I would say the tasks you are looking for is not possible at this moment. The read lenghts are not information that is usually thought to be important/essential to warrant a special searchable field.

Also I think that the concept of fixed read length is just a limitation/characteristic of a sequencing technology rather than being permanent attribute that characterizes all data.

Since the ftp site in the 1000 genomes is accessible remotely via wget, you could query the second line of each file and count the number of characters like this:

wget -qO- ftp://ftp.1000genomes.ebi.ac.uk/vol1/ftp/sequence.index | awk '{print $1}' | \
grep -v FASTQ_FILE | head | while read fe; do echo $file; wget \
-qO- ftp://ftp.1000genomes.ebi.ac.uk/vol1/ftp/$file | gunzip -c | head -n 2 | \
tail -n 1 | wc -c; done

data/NA19238/sequence_read/ERR000018.filt.fastq.gz
37
data/NA19238/sequence_read/ERR000019.filt.fastq.gz
37
data/NA19240/sequence_read/ERR000020.filt.fastq.gz
37
data/NA19240/sequence_read/ERR000020_1.filt.fastq.gz
37
data/NA19240/sequence_read/ERR000020_2.filt.fastq.gz
37
data/NA19238/sequence_read/ERR000021.filt.fastq.gz
37
data/NA19238/sequence_read/ERR000022.filt.fastq.gz
37
data/NA19238/sequence_read/ERR000023.filt.fastq.gz
41
data/NA19238/sequence_read/ERR000024.filt.fastq.gz
37
[...]