Look at it this way:
If they keep the rollout going, there's lots of upside (the success cases), but lots of downside (the failures), and a likely increase in support calls because of the failures.
If they stop the rollout, there is no more upside, but no more downside either.
In this case, the extra upside doesn't bring them any extra income. However, the downside just adds costs. And, if it happens enough, it overwhelms the carefully-dimensioned call-centre staff rotas (of the ISPs), leading to unhappy customers who have real problems.
Under such circumstances, I could see some desire to minimise the downsides, temporarily, even if that curtailed the upsides too. If the problem became really bad, I could see someone demanding the change be backed out. I have, in the past, spent an hour at 6am backing out a change we'd just spent the night working on ... because it wasn't working quite well enough. Telcos tend to be Conservative with a capital C.
But ... none of us know for sure. It is all speculation about what might be happening.