The Newton had several problems. The first it was intended to be an entirely new platform distinct from the Mac. There was zero overlap between the runtime environments of the two platforms. Nothing you made on a Mac was ever going to run on a Newton. This would stretch already thin third party developers even thinner.
The second major problem is it had cheaper competition with a better market fit in the form of Palm. A Palm Pilot was half the price of a MessagePad and did most of the same tasks. It also actually fit in a pocket which meant it could be carried around and used as intended.
A third problem was its OS was an older model lacking memory protection and preemption. By 1997 it was clear that multitasking protected memory OSes were the future if for no other reason increased stability in day to day operations. Rebuilding the NewtonOS with those features would be a major project.
The MessagePads were bulky and expensive. They were too big to fit in a pocket meaning the only way to carry them was a bespoke case or a briefcase. They weren't that capable so a true road warrior worker was just going to get a laptop. Their target market was John Sculley, executives that didn't want to tote around even bulkier laptops.
The Newton didn't make a lot of sense as a product and killing it off with the rest of the Apple peripherals made complete business sense.
Oh, and the genesis of what we now call the ARM computer architecture was the Newton project.
I was rooting for the Newton but at the same time, I found myself mostly using my Palm Pilot (and later my Handera 330) while my MessagePad sat on my desk unused.
When HyperCard was introduced Steve Jobs didn't work at Apple. When he came back, he killed it along with many other projects he had no role in, like the Newton.