Blender. I set up E R T for move/scale/rotate and 1 2 3 for the navigation, which is how I did it in C4D. (Chad Ashley's video on setting up preferences, plus a couple more recent ones, were helpful). Rotation is different as you have to immediately specify the axis with a second key stroke, otherwise it free floats, though I might be missing a trick there.
Courses are cheaper. I'm seeing fully fleshed out courses on various topics going for $20 or $30. The CGCookie guys did an excellent looking bundle of several redone courses for 4.2, and the whole batch of them is $90 total. There's more training. C4D has the Maxon Training Team, a few guys like Polygon Pen making modelling videos when he can, Chris at RL, Tim Clapham doing a good Hello Luxx course every six months, various YouTubers, and a handful of courses popping up on Udemy or Skillshare or whatever. That's all good and useful, but I'd say Blender has at least 5 x the quantity of courses coming out, or higher.
The annual BCON list of presentations just had two different talks on improving the UI, one with Andrew Price making some smart points, the other with five guys including Pablo Vasquez discussing how they'd just spent a year working on UI improvement, and were about to do the same thing for the foreseeable future. C4D has better UI but they showed the before and after for the Blender colour picker - a minor thing in the scene of things, I know - and the before had a jankiness that I found off putting, and the after had the intuitive layout that I associate with C4D. C4D is still even higher here but at a certain point other elements become an elephant in the room beyond how super polished you can make the UI, and C4D has sat rigid in that territory for quite a while.
I don't view Blender as 'better', I just view Blender as different, but its difference now strikes me as interesting and intriguing rather than off putting.
Even though C4D has the word 'Cinema' in the title, I suspect I was gaslighting myself into believing that C4D was an all-purposes app and it would settle into being fully featured for generalist use. I no longer believe this to be the case. I think C4D is certainly expanding its toolset, but I think it's doing this under the guidelines and purpose of using these tools for mograph above all else.
The above isn't an exhaustive list at all, and I'm not closing the door on C4D, which I still like a lot, but if I sit down and try and make a short film over the next year and change with the tools available both, I'll get further ahead with it in Blender than I will in C4D, despite the improvements finally coming into the latter.
I've been interested in learning nodal workflows for a while. Nodes landed in C4D more than six years ago, but what can you do with them yet? And how widespread is the training for it? Meanwhile the Blender Geometry Nodes developers and users are just getting on with things, and still adding functionality and improvements.
In the Blender UI presentations they made it clear they were aware of their shortcomings in usability and design and were dedicated to improving things. The jankiness in Blender is going away with each release, and the functionality is getting stronger. Blender 4.3 comes out Tuesday next week.
I'll follow what C4D brings to the table in the near future with some interest, but it's an interest that sits just below my interest in learning how to settle into Blender, and I think this is a healthy place to be, as I don't need to wait for important feature improvements, or worry for now about budgeting the cost of the app. I'll stop this ramble here for now.