I've looked at that one and similar before, that's Java based so nice and portable, and has changes as recently as 4 months ago so it's not an abandoned project.
The stated goal of that projects is:
"WyldCard strives to offer a high-fidelity reproduction of Apple's original HyperCard,
not a modernization of it:"
Which IMO is somewhat pointless. Why do a 'remastered" edition of HyperCard and NOT at least try to add in all of things HC need XCMDs/XFNCs to do, like color graphics for starters?
A project I'd be more inclined to work on would be one that takes that base (HC 2.x JAVA Port), modernizes the xTalk (a must have 'null' char support to work with binary files for example, HC needed an external just to do that), perhaps combining it with the OpenXION interpreter engine, which is also written in JAVA, and then improve on the combined thing, adding in more capabilities for working with media, images, sounds, etc. Java based is good in that they can run on anything with a JavaVM, or using GraalVM could be pre-compiled into 'native' architecture binaries. A high degree of portability is something we definitely want!
Feature-wise StackSmith is closer to what I've had in mind in a FOSS xTalk, which its xTalk ('Hammer') interpreter is written in C++ (its UI uses some Objective C). But if the FOSS xTalk 'Community' is going with a C++ engine, then why not just stick with the modern (language-wise at least), expanded and fairly stable one that we've already got?