Those other projects featuring the Tiny are very interesting, however they do point to one thing, in my mind at least, is that is best to have a dedicated Arduino style CPU doing the receiving of the 433MHz. I have yet to see a good, simple project that clearly shows how to directly intercept 433MHz with a RPi.
Hence I think using a Tiny as a pre-processor so to speak is probably a good idea. I have found problems with signal strength as well. Most times though, the "Bad reception" has been because of flaky code, rather than RF problems. However to allay my suspicions I have experimented with what would be the "best" antenna and found this solution to be particularly effective. This is a photo of the setup before I went the next step. I have added to it graphically to explain the setup.
The first experiment was to have a length of TV coaxial cable with the shield stripped back for 17cm at the other end from the Tx or Rx (Yellow line). This was fairly good, but I found when I added a matching length of hookup wire to (where I had stripped off the shielding) the shielding, (Green) the performance really moved up a notch or two. The extra RF in/out helped me debug a few difficult problems as I the program was working OK with a strong signal, and it least gave me a lead where to correct the program.
The core of the coax (Yellow) and the balanced other half of it (Green) makes a simple dipole and connecting it to the earth (the one alongside the signal) and signal on the board made a really huge difference. The Arduino that I use 24/7 to gather the weather data off the Oregon Sci sensors is fairly well central to all three. However my development computer is tucked away in a small room with no windows and brick on two sides. Using this antenna was really helpful.
Cheers, Rob