Looking at write-blocking devices I've notice a large variety of specialized hardware based on different source connector types, including USB-to-USB write blockers. It seems as though it should be sufficient to use the latter device paired with simple *-to-USB adapters to handle the variety of connectors (saving hundreds of dollars in the process) so long as these adapters are not capable of writing to the data source by themselves. Does anyone have any additional information as to whether this is correct and considered a viable method of copying data sources universally, or are there operations in these adapters that could potentially taint the source data that makes them not an option?
The safest option is to use a write blocker as close to the device being made read-only as possible. So, for example, if you're accessing a SATA drive over a USB-to-SATA adapter, use a SATA write blocker if at all possible.
Ultimately, this has to do with the fact that there are usually secondary command sets used for tunnelling commands through hardware protocol adapters. The write blocker may not understand the command sets used to speak with the device on the other side of the adapter. Using USB-to-SATA adapters as an example, you may see any of at least three command sets used, classic UMS, UAS, or SAT. Pretty much all USB write blockers will handle UMS correctly, but only some handle UAS and most don't completely handle SAT (because that would require them to completely handle ATA commands too).
For the good quality write blockers, this ends up resulting in the adapter just not working without any side effects. For the bad ones, it can let writes through, or possibly even contaminate the source device without any write commands being issued (because the write blocker may mangle the outbound command). The problem here is that it's very difficult to be certain whether a given write blocker falls into the first type or the second type.