I notice everyone has their personal preferences (not surprising)... eg 0.8/0.9mm solder wire vs >0.5mm etc. I figure the MG Chem 0.635mm is a nice midway?
This "no clean" stuff, should I be thinking about that? I was told by someone there can be some issues with its wetting properties... or something...
Also the MG Chemicals 63/37 I'm looking at has 2.2% RA flux. Most recommends 3.3% flux? I don't want to have to mess around with a flux pencil on every solder joint I do...?
As
RobertHolcombe stated, diameter choice is primarily determined by the joints to be made in order to complete the joint. The dwell time/time on the joint should be kept to ~2 seconds max as a general rule (for both PTH & SMD); you can get away with 3s on PTH with large lead diameters or solder tabs for example. This is to prevent damage to both the components as well as the PCB itself (lifted pads or scorched FR4 material). Personal preference can be a factor to some extent, but it's certainly not the primary determinant.
Now for a couple of examples as to appropriate diameters. You wouldn't want to use .032" on say 0102 passive components (.010" would be appropriate), as the solder diameter is larger than the component. Conversely, you wouldn't want to use .010" to tin large gauge wire either (i.e. .062" would be more appropriate). You can of course wrap a couple or three turns on your fingers, then twist it up to produce a larger diameter solder wire for joints that require it, but you wouldn't want to do this with .010". FWIW, I find both .025" & .032" good solutions for working with both PTH & SMD; but as you mentioned you plan to stick to PTH, .032" is better suited to your needs (or the .035" Multicore that comes with the Hakko). .032" also works well for larger SMD stuff as well as drag soldering. If you were going to do a lot of smaller SMD however, then the .025" or even .020" would be more appropriate.
No clean formulations tend to have a shorter working time as they usually require higher temperatures before they become active & clean off the oxidation. It also tends to be harder to clean, particularly the versions that use synthetic resins (others use modified rosin, which is easier to clean IME). But it works well, and can be left on the boards. Do note it can be a bit tougher to figure out the level of activity (read the data sheet carefully). Rosin based fluxes are easier to figure out the activity (R, RMA, RA; from least to most active).
As per the flux content, 3.3% is ideal IMHO, as there's less of a chance of getting a non-wetted joint. Particularly when you may be working on something with 20+ years of oxidation vs. solder with a lower flux content or flux with an insufficient level of activity (2.2% is still quite usable). As good solder isn't exactly cheap, and is a particular issue for hobbyists, this is where the single roll of tin/lead RA flux solder wire @ 3.3% by weight originates.
But keeping additional flux on hand is still a good idea regardless, especially for drag soldering SMD IC's (so many DIP parts are disappearing). Fortunately, there's a less expensive option for this than the disposable pens; just get a bottle of
MG Chemicals 835 RA flux, and dispense it however you wish. For example, you could dip a small brush in say a thimble glued to a bit of scrap PCB or wood, use a refillable pen, refillable brush pen (my favorite), needle bottle, nail polish bottle, .... You'll figure out when you do and don't need it as you gain experience.
Also as mentioned,
get quality supplies. The upfront costs are a bit more, but it's actually cheaper in the long run (less damage to PCB's as you're not cooking them for seconds on end trying to get the cheap crap to work), and a LOT less aggravation.