Are we discussing direct connectivity between 15169/174?
or via a third party(ies)? I ask, because i looks like RouteViews has, in:

http://archive.routeviews.org/bgpdata/2020.07/UPDATES/updates.20200715.0345.bz2

at last this bit of clues:
BGP4MP_ET|1594784883.404691|A|91.218.184.60|49788|35.213.0.0/17|49788 174 6453 15169|IGP|91.218.184.60|0|0|174:21100 174:22010|NAG||
BGP4MP_ET|1594784883.404691|A|91.218.184.60|49788|35.216.0.0/17|49788 174 6453 15169|IGP|91.218.184.60|0|0|174:21100 174:22010|NAG||
BGP4MP_ET|1594784883.404691|A|91.218.184.60|49788|35.206.4.0/24|49788 174 6453 15169|IGP|91.218.184.60|0|0|174:21100 174:22010|NAG||
BGP4MP_ET|1594784912.621864|A|195.208.112.161|3277|35.206.4.0/24|3277 3267 174 6453 15169|IGP|195.208.112.161|0|0|3277:39710|NAG||
BGP4MP_ET|1594784912.621864|A|195.208.112.161|3277|35.213.0.0/17|3277 3267 174 6453 15169|IGP|195.208.112.161|0|0|3277:39710|NAG||
BGP4MP_ET|1594784912.621864|A|195.208.112.161|3277|35.216.0.0/17|3277 3267 174 6453 15169|IGP|195.208.112.161|0|0|3277:39710|NAG||

I didn't go digging much more than that, but I bet looking at each day before the 15th July 2020 might tell you when routes started to show up with that type of path.

So, 'since last july' I guess is close to the place where 174 started seeing 15169 across alternate paths.

On Wed, Jun 8, 2022 at 9:08 PM Jay Hennigan <jay@west.net> wrote:

On 6/8/22 15:41, David Hubbard wrote:
> Appears HE (ASN6939) is still
> unreachable though… I feel like less entities are single homed to HE,
> but it would still be a calculated risk.

HE is the 800-pound gorilla of IPv6. I would be leery of a carrier
without reachability to AS6939.

--
Jay Hennigan - jay@west.net
Network Engineering - CCIE #7880
503 897-8550 - WB6RDV