IPv6, tunnels and TCP MSS

Some time ago I started to notice some strange “tls timeouts” to some sites hosted at AWS or e.g. GitHub when connection is done over IPv6.
I didn’t pay attention to that for a while, as those were rare occasions, but recently got curious and tried to debug.
To my surprise, I’ve noticed that packets over tunnel are sent with TCP MSS = 1440, even were it should be “clamp-tcp-mss” enabled by default on 6to4 interfaces. (according to https://help.mikrotik.com/docs/display/ROS/6to4 )
Anyone notices same behaviour? Something got broken in v7.x related to clamp-tcp-mss?
I fixed of course by adding line to ipv6/mangle, but it looks like clamp-tcp-mss option in 6to4 interfaces doesn’t have any effect in 7.9.x:

/ipv6 firewall mangle add action=change-mss chain=postrouting new-mss=clamp-to-pmtu out-interface=he protocol=tcp tcp-flags=syn

maybe the chain should use forward ?

/ip firewall mangle
add action=change-mss chain=forward new-mss=clamp-to-pmtu out-interface-list=\
    WAN passthrough=yes protocol=tcp tcp-flags=syn

/ipv6 firewall mangle
add action=change-mss chain=forward comment="fix MTU, make HTTPS happy" \
    new-mss=clamp-to-pmtu passthrough=yes protocol=tcp tcp-flags=syn

in that particular case, the chain forward or postrouting doesn’t matter that much. The post was more about non-functional parameter clamp-tcp-mss=yes on 6to4 interface. if it is not something that is not supposed to inject automatic mangle rules, then probably it should be documented somehow?