Now that tacit has clarified his post and we know he was talking about Safari (exclusively) on a 64-bit capable Mac...
1. A Mac running in 64-bit mode will perform (generally, at the least) better than one running in 32-bit mode...even with no more than 4Gb RAM.
2. > So our readings reflect coding (Right word?), rather than functionality, differences, i.e. the fact that our apps appear to be running in 64-bit mode is no more than an illusion as far as performance goes.
> If by "coding" you mean that data storage is allocated in 64 bit "words" rather than 32 bit "word" where a "word" is an addressable unit of virtual or real memory that is correct. However, since my test results show an actual improvement in performance running in full 64 bit mode I have to question your use of the word illusion. It is pretty substantial to be called an illution.
Apples and oranges here, because in this instance we were talking about apps portrayed by Activity Monitor as running in 64-bit mode on Macs incapable of running in 64-bit mode.
That's the part that I suggested might be an illusion. (Since the only 64-bit apps I've run across so far are Timer Utility and SlimBatteryMonitor I've no real basis for comparison.)
3. > Whether the additional memory resources are needlessly hogging system resources or not is open to question and experimentation. Until tests are conducted, on an app by app basis, that is pure speculation. It may well be the apps run faster and/or more efficiently in one mode or the other.
Again, I was talking about apps that turn up "running in 64-bit mode" on systems that cannot run in 64-bit mode; perhaps I'll be able to get a better insight into the matter as more complex apps turn up with 64-bit capability and I can do some real experimentation.
I hope I've got this all straight, now.