All things in life are a trade off. All antennas have good points and bad points. A Yagi can have great amplification, but can be limited to a very narrow viewing field. A Quadrafilar Helical antenna has good omnidirectional reception but needs amplification. Each is a trade off. The patch antenna has a reception lobe front and back. The one at the back is 10 dB down on the front. That is 1/10th of the signal is received from the back. And 30dB down at some of the holes in the reception lobes.
Most Satellite phones, and GPS is just a satellite connection, use Quadrafilar Helical antennas.
They are certainly worthy of testing side by side with a patch antenna in a real rocket world to see which provides the best real world performance. And I'm surprised the test has not been done or offered as an option.
I did the patch/Quadrafilar testing years ago with a handheld patch antenna mapping GPS and a Quadrafilar one. Both units had signal strength meters that gave the incoming signal strength. Outdoors both worked fine. The Quadrafilar one showed better signal strength and better limit of error circle than the patch one but both were fine for outdoor hiking.
The Quadrafilar was better at hiking in woods and would get a fix from my all-seasons room whereas the patch could not. Either type used as a launch base GPS is fine for rocket tracking as we're outdoors on an open range. The Garmin 60Cs and CsX can be plugged into a D72A or D74A and have a "map-in-the-hand" for APRS (like Beeline or Byonics) rocket tracking. That's great because it can show one obstructions one has to maneuver around to reach the final lie of the rocket.
On Yagi antennas. The beamwidth on the 2 meter (144Mhz) and 70cm (400Mhz) bands are sufficiently wide enough to be able to hold the downlink in flight by pointing the Yagi roughly in the direction of the rocket. On 900Mhz I've been told that the beamwidth of a 33cm (900Mhz) something like 8 element Yagi is too narrow to be able to point reliably at a rocket in flight. I don't know as I've never tried to test it in that fashion. I did read where a person used a large Yagi on 900Mhz to track a rocket in flight but I don't know how aggressive a flight it was.
I've used the stock antennas on 900Mhz trackers for inflight tracking and then switch to a 900Mhz Yagi after the rocket is down. A 900Mhz Yagi does definitely increase the ground footprint with the rocket on the ground. Heck, it's not moving that fast anymore so easy to point at! Would be helpful with a far ranging flight with a 100mW, 900Mhz tracker.
This is an example of a 900Mhz patch antenna:
eBay (ebaydesc.com) If flying very aggressively with a low powered 900Mhz tracker, holding the patch antenna in the general direction of the in-flight rocket allows the best reception and recovery of the NMEA sentences. I've achieved the best decoding of the position data even on flights that are not that aggressive. I drop the patch when the rocket is down and switch to a Yagi. I have mine mounted on a 12 foot fiberglass pole that incidentally was a Shakespeare outdoor antenna for the very early cordless phones on 47Mhz. I used it in 1979 for that purpose. Was fun to take and make phone calls 3 blocks from my apartment. (That was before cellphones mind you.) Now it's repurposed to hold my 900Mhz patch antenna.
Once one GPS tracks a rocket, they'll be hooked walking right up to their rocket. Especially if the flight is totally "sight unseen". I've seen folks demeanor sink tremendously when they overpowered their rocket and it was a totally sight unseen flight. If there was no onboard tracker, good luck finding it unless it's a large launch and someone else lucks upon the downed rocket. If RDF was used one has to start cruising around and see if they get lucky to pick up a signal on the downed rocket. Again, if it's a totally sight unseen flight, it can be a crapshoot for one to stumble into the ground footprint of the RDF tracker. Even a glimpse of an RDF rocket coming in under main, even at a distance can help immensely with RDF recovery. One can get a general direction to travel and are very likely going to get within the ground footprint of the RDF tracker. If in vegetation, they'll find it.
As an aside, a ground mapping GPS that has a "Sight 'n Go" feature can help immensely. Some Garmin models have it. One sights the rocket along an axis on the GPS as it is getting close to the horizon, push a button and the GPS locks a bearing arrow on that direction. Great if one has to go around obstacles. I've used it alone with model rockets that were lost in tall grass! I sight the rocket just before it touches down and walk the line. Sometimes I'm surprised how far I have to walk but I've never lost a modroc using sight n' go. In tall grass, the parachute usually stays on top of the grass with the rest of the rocket underneath. Makes it easier to find. Even if one is never going to do Rf or GPS tracking. A sight 'n go capable mapping GPS is good insurance to maximize recovery of model rockets that land in deep grass or a distance away. Just shoot the line and keep walking. (As long as it's not a totally sight unseen flight mind you.)
With GPS, one has a readout of where the rocket is. Doesn't matter if the winds aloft cause the rocket to drift under drogue 180 degrees opposite of the ground wind direction. That can really confuse people if they are trying to get a visual. GPS tracking can cue where to look to attempt a visual on descent.
I've tracked other people's rockets and witness the 180 degree switch on the mapping program in realtime. Folks are looking the wrong way to try to get a visual on the rocket and I have to yell and point, "Look this way, it switched direction." That is always nice to have if the descent can be seen. If the rocket lands in vegetation, if there is a noisemaker on the harness even if the flight was totally sight unseen, the flier is going to get their rocket back as the GPS will get them up close. The last known position with the rocket in the air just before touchdown is more than likely going to get one within the ground footprint of the tracker for the final lie if the rocket isn't within sight.
GPS tracking even assists with core samples. I had one with a small fiberglass single deploy rocket. Put too much grease on an O ring and some blocked the "blow hole" for the 4F ejection charge. Went totally sight unseen and received one position downrange. I walked out there and there was just the fincan sticking up.
Dug it out, new nosecone, new Eggfinder and good to go. Got the Eggfinder on holiday sale so I think I only lost like $50.00. That was easier to swallow and I got the rocket and the motor casing back. Better than losing the $234.00 Beeline GPS I had on a apogee only, no drogue flight where the drogue didn't blow.
Main chute had a high speed deployment and the harness broke. Interestingly the chute was recovered and didn't show any signs of stress! Got my cataracts fixed shortly thereafter as I missed putting in one leg of the ematch on the drogue. The altimeter I used (an early model Raven) beeped continuity the same if one ematch or two ematches had continuity. Which means it beeped the same with one proper ematch connection and if the second ematch didn't have continuity, it still beeped the same.
Most altimeters have beep patterns that indicate continuity on both channels and if one channel has no continuity, the beep pattern will cue the flier there is a problem and which channel it's on.
I believe later Ravens had this issue corrected.
As mentioned once one tries GPS tracking, they'll be hooked once they have their first totally sight unseen or recovery from deep vegetation flight.
Kurt Savegnago