Lag is ruining KSP – any ways to improve performance

I think the simplest way to do that (and that's how I usually do) is to use MechJeb with the "land at target" function from orbit. Usually, if your lander is well balanced and not too creative with staging, is quite accurate. One time I actually landed on a previous lander while AFK.

In any case, you can use MechJeb to time the deorbit burn, and then fine tune the landing manually.

If you really don't want to use autopilot, yes, maneuvers nodes don't account for atmosphere, so on Kerbin aim to overshoot your target. Actually, you have try to overshoot your target in any case, because if you don't want to land "Jebediah Kerman style" you're going to brake hard to kill your lateral (first) and vertical velocity before landing.

In Kerbal Space Program, Science depends on the location and the type of experiment.

Altitude Corridor

In detail, for Kerbin, this makes five altitude corridors (with some caveats):

• On the Ground (0m)
• Lower Atmosphere (≤ 18km)
• Upper Atmosphere (≤ 69km)
• Near Space (≤ 250km)
• Outer Space (> 250km)

Biomes

On top of that, Kerbin knows various biomes:

• Grasslands
• Highlands
• Mountains
• Deserts
• Badlands
• Tundra
• Ice Caps
• Water
• Shores

Additionally, the Space Center also has various mini-biomes which return separate results, but only when on the ground (above ground, the whole space center counts as "Shores"):

• Administration
• Astronaut Complex
• Crawlerway
• Flag Pole
• LaunchPad
• Mission Control
• R&D
• Runway
• SPH
• Tracking Station
• VAB

Experiments

Depending on the experiment, some of these altitude corridors have per-biome results, others only have one result for the entire corridor.

In particular:

The Surface Sample is a biome-dependant surface-only experiment.

The EVA Report is biome-dependant on the surface, in the lower atmosphere and in near space; it is global everywhere else.

The Crew Report is biome-dependant on the surface and in the lower atmosphere; it is global everywhere else.

The Mystery Goo Observation (available through the Mystery Goo™ Containment Unit from the Basic Rocketry tech node) is biome-dependant on the surface; it is global everywhere else.

Storing Experiments

There can only be one crew report per part that produces a crew report (that is, per capsule), which is stored in the part. It can be transmitted, in which case it makes space for a new crew report.

Both the EVA report and the Surface Sample experiments can be stored in the capsule as well, with each combination of altitude corridor and biome once. So, you can store a surface sample from every biome, and EVA reports from every biome all at once.

Transmitting Results

Results can either be recovered by landing the craft on Kerbin and then recovering it from the Tracking Station (or from the craft view, where it's hidden at the top of the screen), or by transmitting them home.

Transmission is hardly effective this early into the tech tree - you lack the electricity to send the results back. Nevertheless, you can - for example - send back one crew report to make space for another, so here's some information about transmission as well:

Every experiment has a transmission efficiency, which reduces the Science yield on transmission (as opposed to recovery).

In the case of the Surface Sample this is 25%, which makes them an undesirable candidate for transmission.

Similarly, the Mystery Goo Observation has only 30% efficiency.

All reports (be that EVA or Crew), however, return a full 100%.