This statement resembles my thinking, too. It’s just such a sequence of unfortunate events which - even if not impossible - seem improbable: incredibly short timeline, physically fit individual/swimmer, broad daylight, no witnesses, quick emergency response, no screams for help reported, no traces on riverbank, extensive search, phone left behind …I agree with the sentiment but I think we still have to appreciate based on the evidence we have the theory that she entered the river an extremely bizarre turn of events. Of course there's reasons why she could have been at the river's edge but then I could equally argue the opposite, why would you go to the water on a cold day like that, there's not much to see, you have a good vantage point of the river from the bench to see if the dog was in or down by the river, the first instinct would be in the open field. What's puzzling for me is when you put it all together, going to the water's edge, falling in, falling in whereby you leave no trace on the bank, falling in but not being able to get out when it was only waist deep at the bank, no sign of the body when the witness reached the bench, very slow flow to the river (the other story was when it was flooding I believe), divers on the scene very quickly, the weir, the shallow rocky part over the weir, dog didn't enter water, no sign of anything on the sonar.
Now all of these things are possible and it's the probably most likely explanation but that sequence of events is very hard to comprehend to the point where we have to give serious consideration to 3rd party involvement.
But perhaps the police have something more to believe differently
I cannot get my head around the phone/work call. Was it a call she was actively participating in or expected to contribute to at any point? If so, it is unlikely she would have moved far away from her phone whilst on the call, even if taking a quick look at the river.