I support pretty much the entire article.
Point by point...
1: ... Yes. Always draw first. Runners are the only ones who can get away with not drawing first, and it's because they have specific reasons to do so (Runs being the primary motivator, as you mentioned). The only Corp I know of that can legitimately draw on last click is Cerebral Imaging. The reason can be explained very simply: You would rather have randomness happen, then mitigate it, instead of taking your actions and then hoping randomness favors you. Thus, you'd rather randomness happen first in your turn (Drawing a card, making a Run on an unknown Asset/Upgrade/ICE or a Central) and then deal with it since it gave you new information you couldn't have had. (You can't know 100% the result of those earlier things).
2: Ehh, sure. I generally do this and support doing this, but I think it generally follows from point 1. It also relates to a central tenet of Netrunner: Have a Plan to Win. You can't just Not Lose. Do you win on R&D runs and lockdown? Win on Remote Camping? Fast Advance? Taxing? Flatline? Every action you take should advance you towards that goal. (This is why there are very few Corp decks that contain 3 win conditions. You can usually advance your Rush and Kill plan at the same time, but it's very difficult to Rush, Kill, and Fast Advance. Unless you're NBN.)
3: Ha, yes. Nothing worse than rezzing a card and realizing that was absolutely the wrong thing to do. Some will go further and say you shouldn't ever look at your ICE on servers that aren't being run. (If I see you check the ICE I'm running, then check the ICE on that remote you put an Agenda into, I'm fairly certain you're calculating costs. If you then don't rez the ICE I'm running, I become fairly certain you can't rez all of your ICE and Score the Agenda.)
However, my counterargument is that if you look at ICE/Unrezzed cards, look at ALL of them. You don't necessarily have to memorize all your facedowns, but if you don't or can't (which is fine!) then if you check some, check all of them to avoid giving away information other than 'Checking to make sure all of these are what I remembered'.
4: Yes. It's so important that I'll often Mandatory the ICE they just saw, draw another card, then install an ICE, saying 'Installing that ICE you just saw.' Whether it was or not, it plants doubt. Absolutely pay attention not just to the information YOU have access to, but the information your Opponent has access to.
5: Yep. I've had it happen for and against. I do want to emphasize that you can usually learn SOMETHING even during games decided by Variance. Here's some examples:
-- Playing Sunny, I go through 30 cards before I find any one of my six drip econ cards. Game is lost. This is due to variance, BUT, I can use this as an example for the 'This deck struggles if it doesn't find Drip Econ early.' Thus, I've learned to install those cards quickly, and mulligan'ing for them is palatable.
-- Playing against a Corp, I quickly find 7 points of agendas through random single accesses on R&D and HQ. Only 9 agendas in the deck, and when I got the one from HQ, it was 1/5 cards in there that do anything when I access. Really lucky win, but I can use it to learn two things: First, "Check Centrals, keep pressure up." Sometimes you actually CAN win a game just by keeping good Central pressure. Second, for the Corp side of things "Don't blindly trust RNG to save you on Central accesses. DO Something about it!"
6: I'll just say Yes to this whole thing, with one Caveat: I used information from opponent's Heap to inform my play... I played out an agenda, it got Inside Job'd. I play out another, to force his hand... And he IJ's again. Surely, with two gone and only 15 cards deep on his deck, the third one will work. So this time I install an agenda I actually care about scoring instead of just using as bait. (I'd used a Hostile Takeover for the first since I wasn't confident it would survive, second was an Oaktown. Third I believe was The Cleaners perhaps. But generally that sort of progression is what I'm going with.) So, of course, he has the third Inside Job. Just because you see 2 of the likely 3 in the Heap, doesn't mean he's any less likely to actually HAVE the third in his Grip already. Counting Influence is a seriously good advice, though.
7: Oh my, yes. Especially with RP or any Glacier. I feel this is the most important lesson you must learn if you want to play: HB EtF or other Never Advance variants... RP Glacier or other Glacier-style... Argus Rush... PE Cambridge... Sunny... Nexus Kate... To put it another way: You don't need all the points, just the 7th. No one will care that you beat them 7-0. You don't NEED to beat them 7-0. You can give up 6 points if it guarantees you the 7th point. (Aka: Let them score those two Improved Protein Sources. Just don't let them have the Clones Aren't Dudes at the same time, k?)