That's a position shared by a lot of people and I don't think it's unreasonable. It's absolutely possible that she harmed those babies and possible that she harmed even more. Staff in a hospital are in a position of trust and it's not really possible to track their every move or monitor their every action.
It's also possible that the babies weren't deliberately harmed at all as was thought at the time. From the other neonatal experts it seems that the standard of care was not good enough at the CoCH. From expert statisticians it also seems that although there was a spike in incidents, this was not outside the bounds of probability.
Personally I think there are some questions about the way that the evidence was presented to the jury - in particular what was left out - and I think there has been new evidence that has come to light since the trial about other similar cases and about the way the hospital was run. I also think that with so many methods of harm suggested that it was inappropriate for the judge to say the method didn't matter, only that they were sure it was deliberate. Having 18 cases would also make the jury think that she must have done something, even if none of the cases individually were particularly strong.
There is no direct evidence that Lucy Letby deliberately harmed any baby. If direct evidence did come to light I would absolutely change my position to definitely guilty. In fact, if there was some definite evidence of deliberate harm - clear differences between the incidents determined to be "suspicious" vs "not suspicious" then I'd say it's much more likely that she's guilty.
Thinking particularly of the xray of baby C that turned out to have been taken before Lucy Letby even met that baby, it's not good enough to present that as evidence of harm and then walk it back to "CPAP belly" when it's found that she couldn't possibly have caused it. If you're absolutely convinced that you have evidence of harm then you need to be looking for another culprit. If there's another possible cause then it's not evidence. You can't imply guilt by saying she was there for every suspicious incident if part of the criteria for determining whether an incident was suspicious is whether or not she was there.
The evidence is all circumstantial. Circumstantial evidence can be enough to convict someone if it's strong enough. What you shouldn't do is start with a culprit and cherry pick evidence to fit that.