Not apples to oranges, but I had much the same experience. Support is A1, undoubtedly. The Enta backend wasn't exactly optimal for me though, and since switching to YouFibre my pings have dropped significantly (7ms to London vs 12ms, and 1-2ms to Manchester vs 15ms due to not routing Liverpool > London > Manchester). Anecdotally, my BQM/latency is much flatter too - no need for SQM and the graph is a flat 7ms day and night no matter what.
It's worth looking into the issue further if you're seeing real world issues arising from it, rather than just the graph not looking as sleek and pretty.
Code:
$ ping -c 4 -i 0.5 lon.speedtest.clouvider.net
PING lon.speedtest.clouvider.net(2a0d:5082:0:4::2 (2a0d:5082:0:4::2)) 56 data bytes
64 bytes from 2a0d:5082:0:4::2 (2a0d:5082:0:4::2): icmp_seq=1 ttl=54 time=7.35 ms
64 bytes from 2a0d:5082:0:4::2 (2a0d:5082:0:4::2): icmp_seq=2 ttl=54 time=7.07 ms
64 bytes from 2a0d:5082:0:4::2 (2a0d:5082:0:4::2): icmp_seq=3 ttl=54 time=7.28 ms
64 bytes from 2a0d:5082:0:4::2 (2a0d:5082:0:4::2): icmp_seq=4 ttl=54 time=7.46 ms
--- lon.speedtest.clouvider.net ping statistics ---
4 packets transmitted, 4 received, 0% packet loss, time 1504ms
rtt min/avg/max/mdev = 7.070/7.289/7.459/0.141 ms
Code:
$ ping -c 4 -i 0.5 dns.quad9.net
PING dns.quad9.net(dns9.quad9.net (2620:fe::9)) 56 data bytes
64 bytes from dns9.quad9.net (2620:fe::9): icmp_seq=1 ttl=59 time=1.64 ms
64 bytes from dns9.quad9.net (2620:fe::9): icmp_seq=2 ttl=59 time=1.74 ms
64 bytes from dns9.quad9.net (2620:fe::9): icmp_seq=3 ttl=59 time=1.74 ms
64 bytes from dns9.quad9.net (2620:fe::9): icmp_seq=4 ttl=59 time=1.58 ms
--- dns.quad9.net ping statistics ---
4 packets transmitted, 4 received, 0% packet loss, time 1504ms
rtt min/avg/max/mdev = 1.578/1.673/1.740/0.068 ms