ZaltysZ

P51 has steerable tail wheel.

Get 64 bit dependency walker: http://www.dependencywalker.com/ Open dcs.exe with it check list of DLLs it depends on. There is a CPU column in middle section, and it should say x64 in every row. If not, you will likely have misplaced 32 bit DLL - such things happen with some old/bugged (not DCS) installers. Also, it will show if you have any missing dependencies (some of them, like API-whatever can be safely ignored though). Work from there.

The problem with your question, documentaries (or TV shows) and interviews is that 109 is presented like it had belonged to very "narrow" series. It had not. 109 series was very broad, spanning over whole WWII and a bit more. There were great performance differences not only between variants, but even between subvariants or even between field modifications (Rüstsätze). BF109G6 in antibomber role was nowhere as scary as clean BF109K4 performance wise, and P51s could meet them both in 1944, that is why all the tales about P51 out performing BF109 must be taken with grain of salt, especially if BF109 is so vaguely described as just 1-oh-9.

Overboost? Low rpm + high boost is bad for engine and normally automatics tries to prevent that by adjusting prop pitch when you operate the throttle. However, prop pitch change is kinda slow on BF, so in certain situations automatics can't follow you fast enough. I.e. if you do a long dive, pick speed up, prop pitch will be adjusted to coarse to load engine and keep it below redline. If you dive with reduced throttle, automatics will try to achieve even coarser pitch and reduced rpm. The problem arises when you bleed energy after such dive quicker when prop pitch can be adjusted. Suddenly engine becomes overloaded, its rpm drop, you firewall the throttle, boost rises, but engine still struggles rpm wise, because prop pitch is still way too coarse. Now you are heavily abusing the engine.

No. You sound like you think that trim tab is additional mini elevator. It is not. At least not in context of whole aircraft. If you want to lower the nose through stick, you move your elevator down, however if you want to do the same through trim tab, you move the tab up. Why? Because it will move the elevator down. It is elevator which changes the attitude of nose, and trim tab just helps in changing the angle of elevator exploiting the leverage. Delete the elevator, and aircraft with trim tab only will be hardly controllable.

It is by design of Sukhoi. Stability system "fakes" forces to give feedback to pilot.

You have to go easier on the stick at the top of the loop. Depending on trim, stick can even be neutral in pitch axis. Otherwise, if you push the plane to max through wide speed range, you will get non round loop as radius decreases with speed. It would be easier if you made a video. Otherwise we will be speculating only on most likely reasons of your issues. BF109 is supposed to be heavy on controls. As speed increases, elevator becomes heavier and heavier. I think at speed of 400km/h your virtual pilot can't event pull stick fully. That means, once you pull your stick over the strength limit for given speed, further input in that axis is ignored until you ease on the stick and go below that limit. I think you initially pull too much, go way over that limit, and then don't see immediate difference by going easier on the stick, because your real stick still stays over the limit. Your easing is ignored until you ease enough to go below strength limit. Once your speed bleeds, forces on elevator decrease, and your virtual pilot becomes able to pull more and more closer to your real stick. At certain speed forces becomes light enough for virtual pilot to pull near critical AoA, and if your real stick is still held pulled more than that you will be moment before shaking starts. Basically, you need to understand how control heaviness is modeled and adapt to it (learn not to pull over the strength limit of the pilot by using visual cues and/or IAS reading) or get FFB stick. Comparing to what red square (CTRL+ENTER) shows to your real stick position should be helpful too. Concept of stall speed is more useful for landing. While doing maneuvers think about angle of attack. Wing stalls once you reach critical angle of attack, and you can do that at speeds higher than stall speed. Slats (front flaps) help with high angles of attack. If they were removed, you would be getting problems at lower angles of attack.

Civil traffic: cars, trucks, trains, buses have been in DCS for long time already. In fact, they were available even in DCS predecessors. Were they removed in 1.5?

Trim tab does not change pitch directly, however it affects the elevator by applying additional force to it. Because stick is connected directly to elevator, additional force is also applied to stick and its hands off position moves off center. At higher speeds this force can become so large, that pilot simply won't have enough strength to move the stick where he needs, if aircraft is trimmed too much in opposite direction. Unless you have FFB stick, there is no intuitive way to feel this force. Simple spring loaded stick will still move full range while tricking you that same happens to elevator. It does not. Check stick in the cockpit or press CTRL+ENTER and check stick position in red square. In high speed flight with nose heavy trim, cockpit stick cannot be pulled much, despite your real stick is pulled fully.

If you don't mean "nice for an aerobatics competition", then maybe this would be good enough:

How it is different in this regard right now? Most HOTAS have multiple HATS on throttle, so it is even possible not only trim while actively maneuvering, but even while operating the throttle. There is no "security" in this weird implementation too.

Probably not. Manufacturer describes it having 9 rotations, so its axis reports absolute position and not the turning rate. Basically you will need a turning rate reporting controller, if you want to trim current implementation of our K4 in not so mind hurting way.

Is there a difference between FFB and non FFB? Does FFB stick want to rest off center by itself in same situation (same speed, trim, no curves)?

"Object" is too vague as it can be 3D model, physic engine object or combination of them. When people speak about smart scaling, they mean rendering of 3D model with applied scale factor. The old school (and accepted) way is to make physics engine influence the graphics engine and not vice versa, so something like visual representation affecting radar is so weird, that I am even inclined to write it off as miscommunication. However, there are more than enough graphical problems related to smart scaling and those make it a no go in visually detailed environment. I.e. objects partly intersecting without collision, almost blending with each other on collision, "wrong" shadows and lights, too dense or too small debris and etc.

What SSD do you have? There are some SSDs which cause stutter in DCS when OS is installed on them no matter where DCS is installed. I think the reason is temporary file location and too slow TRIM implementation on these SSDs. Try this for a test: http://forums.eagle.ru/showpost.php?p=1951223&postcount=38