I'm curious to hear why a breaking change like that did not warrant a major version change? Not really a critique - I'm not a user of Chicken Scheme, so I assume that it's no problem for the people who actually use it.
(There's actually a historical headache in Schemes due to not standardizing keyword objects over a decade ago, and that headache gives implementations some leeway, IMHO. Lots of people really wanted keywords, enough to implement them without a standard or even coordination. So there's a weird messy set of semi-compatible and unnecessarily complex (and sometimes visually ugly) ways of doing keywords in various Schemes. Given that situation, it's difficult to "break" anything further, regarding interpretation of any syntax that looks like keywords. For example, as a Scheme programmer, you probably wouldn't have written any code in the last decade in which you require `:foo` to read as a symbol, at least not without knowing that's non-portable and something you'll have a good chance of needing to change later.)