Why not add support for the qcow2 format used by qemu/kvm in the image generation

This should be able to provide convenience for qemu/kvm users.

About kvm:

I have no problem at all to use the official images as-is with qemu. Why is qcow2 needed?

The VMDK image can of course be booted, but the snapshot function of qemu/kvm can only be used in qcow2.

$ wget https://downloads.openwrt.org/snapshots/targets/x86/64/openwrt-x86-64-generic-squashfs-combined-efi.img.gz
$ gunzip openwrt-x86-64-generic-squashfs-combined-efi.img.gz
$ qemu-img convert openwrt-x86-64-generic-squashfs-combined-efi.img -O qcow2 openwrt-x86-64-generic-squashfs-combined-efi.qcow2.img

…not really more difficult than gunzip.

$ qemu-img info openwrt-x86-64-generic-squashfs-combined-efi.qcow2.img 
image: openwrt-x86-64-generic-squashfs-combined-efi.qcow2.img
file format: qcow2
virtual size: 120 MiB (126123520 bytes)
disk size: 14.8 MiB
cluster_size: 65536
Format specific information:
    compat: 1.1
    compression type: zlib
    lazy refcounts: false
    refcount bits: 16
    corrupt: false
    extended l2: false

Of course I know qemu-img, but why do other virtual machine software users have native virtual disk mirroring support, while qemu/kvm users don't? Is it because qemu provides a conversion tool, is this an unfair treatment for qemu/kvm users?

Especially when vhdx support is added to the mainline code, but qcow2 support has not arrived, this frustration is particularly obvious.