I lean towards wing tip sensors
As soon as a thermal starts shedding swirling parcels of air, you will detect nothing but "false positives" if you are envisioning the Snooper output as somehow indicating the shape of that thermal. What the Snooper output is telling you will be that there is a thermal nearby and that you should initiate a search pattern. Because of this, there is no advantage to wingtip sensors.
Knowing whether to turn right or left is not as important as knowing
A) if you are in a rising (or descending) and swirling parcel of air shed by an ascending thermal (near the thermal: Snooper noisy, vario indicating some lift) or
B) within the thermal itself (Snooper quiet, vario screaming).
You may be thinking in terms of
rate of change in pressure (barometric variometer). Thinking along those lines leads to rationalizing that a differential between wingtip sensors is presenting information when it is simply noise.
A Thermal Snooper, however, is providing a
rate of change in temperature (temperature variometer) along a line (flight path). This should not be confused with a thermometer or change-in-temperature instrument (temperature "altimeter"). Such confusion might lead one to think placing very sensitive thermometers on each wingtip might provide useful information when 30 or 40 feet of separation is insufficient.
What is the minimum separation needed to generate information in an ideal environment, rather than be overwhelmed with noise? Would simple wingtip temperature sensors (thermometers) work on a 4-foot span model? If not, why then assume they would work on a 30-foot span? I contend the minimum useful distance of separation for any soaring aircraft will be a function of its turning radius and likely much more than its wingspan. But if you then add turbulence generated by the thermal, including temperature turbulence generated by swirling parcels of warm air sloughed from the thermal and mixed with swirling parcels of cooler air from the boundary, it is likely you will end up with nothing useful.
Now let's analyze the preconception that wingtip rate-of-change temperature sensors (Snoopers, not thermometers) are a practical approach to determining the location of a thermal. These wingtip sensors are imagined to provide a differential over a short and set distance perpendicular to the flight path. Furthermore, they are assumed to provide more useful information than a single central sensor located only 15 feet between them.
Compare this to a single sensor. At 24 mph you will be travelling slightly more than 35 fps, about a wingspan. If you take a measurement, then count one second and take another, the result is the same snapshot rotated horizontally by 90 degrees (and skewed by one second whereas the wingtip sensor measurement would be simultaneous). The preconception would suggest that the single central sensor has less value in making a turning judgement than wingspan sensors. The preconception also suggests the value of a temperature variometer is in providing information as to turn right or left.
But this is not what a rate-of-change-in-temperature instrument does. A pilot is much more likely to encounter parcels of tumbling air near thermals rather than enter a thermal by blind luck. The Thermal Snooper tells you a thermal is nearby. That is its purpose. (Unless you actually approach the thermal core by blind luck, all an ideal wingtip sensor equipped instrument could possibly do is tell you if the parcel is rotating clockwise or counterclockwise along your flight path axis - which is irrelevant.) You execute a search pattern as illustrated in my last post. You have a 50/50 chance of turning in the correct direction. If the thermal is there, you will find it with your altitude variometer and determine you have centered the core when the temperature rate-of-change variometer (Thermal Snooper) falls silent.
People who choose to disregard the Sloughing Thermal Hypothesis, choosing the classical thermal model instead, are not as proficient in using the Thermal Snooper because they keep thinking they are encountering little thermals.
Home
as many have thought 
)