The main benefit in killing PPC support would be the reduced filesize; and the lack of any sort of commitment by me to fix bugs with it. In practice, I haven't been doing much PPC-related for a while, anyhow, but dropping it would make that official.
There are also issues with other parts of the toolchain losing support for PPC macs. For example, I don't think the text editors I'm considering for inclusion in future releases can be run on the mac. Similarly, the latest version of the Xcode toolchain will drop support for PowerPC development entirely.
When I do a rebuild of Ren'Py, the mac version finishes last, since it has to build two versions of the code.
Finally, the Mac App Store doesn't accept applications that include PowerPC code in them. I'm not committing to support that at any time in the future, but it's at least a consideration.
Why would I buy a newer Mac? My laptop still runs. It plays games. It plays videos. It runs text editors. It accesses the internet.
Does it? I mean, sure it can run the games you bought a while back, and Ren'Py games - but how many games are still being released with PPC support. How many text editors? It might be able to access the Internet, but it can't run supported versions of Chrome or Firefox, and support for Safari (along with security support) will be dropped this fall.
From a testing perspective, I'll point out that you're not actually testing the code that your users are running. It's close, but so is (say) the Linux version. (You _are_ testing things like integration with the finder, but not with the newest versions.)
The thing here is that we have a dead platform - one for which no new hardware will be released, ever. How dead does a platform have to be for me to drop support for it, since there will be operational PPC macs two decades from now? I mean, I have a working Apple ][e in my parents' basement.
The last PowerPC mac was discontinued in August 2006.
I'm curious as to if people have a sense for how many end users are still using Power Macs. I really don't want to disenfranchise people like Papillon or her users, but at some point, I think it's time to end support for the platform.