There's the assumption made that "old" (4-years) = worse than anything new out there. That's a false assumption. If a company gets it right from the get-go, it doesn't matter how old the technology is, it's still better than a lot of the newer offerings out there. This is especially true in the plug-in world. People still use "old" McDSP plugins (admittedly, they have had their updates) over newer offerings. There are many reasons for this, but one of them is surely: "the plugin is just that good".
Now are the POD HD models "that good"? I'd say "good enough for me." With regards to amp sims, you have to think about the process of modelling. Do you want to model everything, point-to-point? But then the programming will be bulky and slow. Do you model "chunks" of the circuit that have no audible effect to slim down the code? Do you do the same with parts of the signal that affect the audible parts? How "precise" do you want your algorithm for a tube or transformer and all its irregularities to be? At what cost?
I think Line 6 has done a good job making the right programming compromises to give us an excellent amp-sims. Every company has to make their own programming compromises, even with newer offerings. I can think of one example of a newer amp-sim plug-in from another company that people love, but to me their plugin obviously makes compromises that don't make their amp sims amp-like (in terms of how each tone control interacts with other parameters for each unique amp, and in terms of how each amp has very, very different clipping characteristics). For some more reading see this article on plugin modelling.
With regards to price, $50 is very, very reasonable. 4 amps and 2 cabs = 6 very unique tone shaping tools. UAD is charging $150 for three pedal sims. Now THAT is ridiculous.
My only bummer is that apparently most of these new "package" sims are older-generation sims. I enjoy the HD models much, much more than the older-generation stuff. It sometimes feels like the clipping in the older generation stuff is too similar amp-to-amp. I'm not sure if I'm right or wrong with that assessment.