In addition to the above GPS issue, for wind vectors to be true and correct, you need the following.
1. OAT and this needs to be verified it is correct, not just on the ground in your hangar, but in the air. I have seen plenty of bad OAT installs being affected by engine heat/exhaust. People put them in places under fairings where they can over read also. Best place is through an inspection cover out on the wing, like beside a pitot tube.
2. Static......not static checked on the ground either. YOU MUST validate the static in the air. This is done by flying a GPS box and calculating the actual TAS and comparing that to your TAS derived from IAS. Or you can run a seperate test boom or drogue to compare static. GPS box is easier. If your TAS is not correct or within a knot or two at the most, then you have no hope of getting a good wind vector. Not to mention you also have a altimetry error and that could be significant and dangerous in extreme cases.
3. Compass, this must be pretty darn close to spot on, obviously a degree or two will only have a minor error but the better this is the better result you get.
4. IAS.....obviously accurate IAS from your instrument test. Dynon have an excellent record of supplying very accurate units, so this will be fine provided your dynamic static is accurate like it is on the ground.
DO NOT ASSUME you have accurate OAT and STATIC, this is the biggest error causes in TAS, and most misunderstood. A large percentage of the experimental fleet have errors ranging from only a few knots up to double digit and 5-10dC in OAT error. 6-9 knots of TAS error could mean 130-150 feet in altitude error. You must get these right, and you must prove they are right. You cannot trust, assume or believe copying others installation is good enough.
Hope that helps.
