I think most people who are questioning the conviction aren't saying "she's definitely 100% innocent" though. Most people are worried that the evidence presented/reported was misleading, rather than deliberately scapegoating.
For example, you mention that the death spike stopped when she stopped working there. Well, they also downgraded the unit from level 2 to level 1 at the same time. Since that time the unit (old and cramped) has been completely rebuilt, which they were fundraising for at the time LL worked there.
To actually determine whether you have a spike in deaths/collapses you can't just look at the raw figures, you need to analyse the data to see if there are additional factors causing it (e.g., greater acuity, more babies, other units closing, increase in problems on the maternity side, staffing problems) and whether you have a genuine outlier. A statistician was originally employed by the police but dropped after the CPS told the police not to pursue it. That was never reported in court.
To even start accusing someone of murder you have to determine whether a murder has been committed. This is the basis of the doubt that a lot of people have. There were other deaths and collapses on the unit during the period. The babies were in NICU/special care for a reason - they were either premature or sick. Saying that they were "perfectly healthy" or that they weren't expected to die when the consultants were only doing ward rounds 2 or 3 times a week instead of the expected 2 times a day means that the signs they were deteriorating could well have been missed.
You would hope that there were clear lines between something that was deliberate harm and something that was down to natural causes or poor care in an overworked, understaffed, underskilled department but I'm not sure this is the case. There were initially at least 28 "suspicious" events identified by Dr Evans. Some of these were later dropped when it was found that LL wasn't on shift so were they suspicious or not? Some of the cases were not thought suspicious at the time but were identified after a trawl through the notes. The defence were not allowed to mention any babies who were not on the indictment list.
The jury were told that the only way the insulin readings for babies F and L were possible was if someone had deliberately given them insulin that they didn't need. The prosecution came up with a theory that insulin had been added to their TPN bags. This was because in one case, although LL was there when the first bag was hung, she was off duty when it was replaced and the dubious readings continued. She would also have had to add insulin to that second bag somehow. The first bag was a prescription one where the pharmacy remove the original seal, add to the bag through a port and replace the cap with a different type of cap which could in theory be removed and replaced by someone with ill intent.
However, the bag wasn't due to be replaced and she would have no way of knowing that it was or which new bag would be used. Also the second bag was a "stock" bag which would still have been in its cellophane wrapper with the original tamper proof seal. Additionally, insulin isn't normally added to TPN bags because it "sticks" to the plastic so you can't be sure what dose you're giving. To be clear, no TPN bags were kept (this did sometimes happen) so there's no evidence that they ever contained insulin and the insulin on the ward wasn't tracked so there's no evidence that there was any missing. There's, I think, 2 other babies who also had anomalous readings from these tests which wasn't considered suspicious.
The "new" evidence from Chase and Shannon relates to whether the readings were actually suspicious in the first place. They concluded that they're not. As I understand it, their withdrawal from the defence team relates to the other possible theories - that the tests themselves were faulty - were very unlikely.