This article sums it up pretty well -- you need to be able to work out the actual word; whole-word approaches don't work well.

https://www.apmreports.org/story/2019/08/22/whats-wrong-how-schools-teach-reading

'...in 1975, Stanovich and a fellow graduate student set out to test the idea in their lab. They recruited readers of various ages and abilities and gave them a series of word-reading tasks. Their hypothesis was that skilled readers rely more on contextual cues to recognize words than poor readers, who probably weren't as good at using context.

They couldn't have been more wrong.

"To our surprise, all of our research results pointed in the opposite direction," Stanovich wrote. "It was the poorer readers, not the more skilled readers, who were more reliant on context to facilitate word recognition."13

The skilled readers could instantly recognize words without relying on context. Other researchers have confirmed these findings with similar experiments. It turns out that the ability to read words in isolation quickly and accurately is the hallmark of being a skilled reader. This is now one of the most consistent and well-replicated findings in all of reading research.'

Eventbrite is set up for this sort of thing -- volunteer signup is a very common use. Plus, free.

I think I'd set up a small demo using your real data and requirements -- its obvious ease of use and lack of cost would make the case pretty easily. (I'm sure that's the case for ten or perhaps fifty other off-the-shelf services too.)

I'm surprised people have been giving Feedly such an easy time over that claim, which to me seems deeply suspect.