Best Type of Proxies for Scraping (specifics)

I am wanting to know a little bit more detail in regards to the best types of proxies I should be using for scraping SE's (with Hrefer, so not only google)

For harvesting, I realize that public proxies are the way to go when dealing with mass. That is pretty common knowledge. But I want to know more.

What else should I be paying most attention to, and what should I not care about?

- Some say 'Anonymous' is not super important (confusing).
- Some say both HTTP & SOCKS work, SOCKS are faster, but harder to find.
- Some say to stay away from China/Pakistan proxies (as they wont harvest)

I realize that I can just do some tests and probably figure this stuff out on my own, and I am, but it would be super appreciated if somebody very knowledgeable about proxies could chime in and help me shortcut my testing

You can use free public proxy but if you want to get a much better speed you must use private proxy. Also you will know that you are always anonymous and you will have a better experience.
Yes, you can use SOCKS proxies.

I didn't put enough emphasis on the 'harvesting MASS links' part.

A lot of this depends on what you're using them for. That will determine what anonimity level you'll need, and that will give you some direction as to what to look for. If L3's(Transparent) proxies will do the job then there's thousands to choose from. If you need to completely hide your IP that will bring less results. For something to start with for bulk links to harvest from go to Google and paste in the following code there and you'll get a ton of results.

You can do a number of things from there to filter/narrow your results down a bit, but that should give you nice list to start with of links you can grab some proxies from.

About your anonymity question, it's pretty simple.

Some proxies add X-ORIGINAL-IP(or something like that) header and then identify you as the source of the request. Thus you get people asking "I used a proxy and google said automated requests anyway". That's right, with that header on, the request came from your IP, not the proxy as far as they care.....

you are right china proxies are not good for scrapping at all. If you are looking for list of public proxies there is a guy here called @proxygo contact him.
Abut the socks, they can not be faster then http/https because the servers load is higher due to encryption, the speed really depends on peering from your ISP to the server network where the proxy is hosted, how far the server where the proxy is hosted is, and so on ...
If you are looking for high speed private proxies that have a UP-Time of 99.99%, take a look at my signature

if you use public proxies, you will have to replace them a couple of
times per day

if you use private / shared proxies provided you keep your requests down
they should last a month, the more aggressive your searches the faster the
proxie is banned