You can claim others are doing the exact same thing as me, but I'm not seeing it at all. At least I provided one usage case where a Fusion Drive is better than a 256Gb SSD.
You are not understanding what I'm saying. It's not about ssd or Fusion Drive being better. It's about how you presented it. You gave a use case that was suited to a particular kind of drive which is what the others are doing as well. To word it differently:
You say: buy Fusion Drive if you don't have much money and/or don't do heavy computing.
Others say: buy ssd if you have the money and/or do heavy computing.
Same difference
In fact, I challenge you to find ONE post where someone is explaining a valid usage case where a 256Gb SSD is better enough to justify additional cost and hassle of an eternal drive. These cases do exist, but I didn't see them mentioned on these forums.
Start reading your own replies (you actually confirm one of those use cases and started complaining that he's doing the same thing as you do) and many other posts on this forum. It has been mentioned several times that ssd's aid in any use case that requires fast io/generates lots of iops. Among them are things like video editing, databases (some of the people here run OS X Server!) and virtualisation (which is sluggish on a Fusion Drive due to data being on the hdd and needing to move to ssd...in blocks thus not the entire virtual drive is being moved). Having multiple drives has been discussed here: you separate OS and apps from your actual data (documents, photos, movies, etc.) which is common practice in IT. The reasons for that are quite simple: it makes recovery really easy and you can use whatever medium is best for the data. Have OS and apps on an ssd so you can quickly start them and put data, like mp3's, that see no benefit from anything faster than an ordinary hdd on an hdd or maybe even a NAS. If your data is more than 3TB then there simply isn't any other choice but to go external.
There are also some that point out the added complexity of having 2 drives (lo and behold, a Fusion Drive is no different than the dual drive setup because it consist of 2 drives!) Fusion Drive is just a software thing (what happens if an OS X update/upgrade messes it up...): it's a volume spanned across an hdd and an ssd where most used data is moved off the hdd to the ssd in blocks. Fusion Drive does perform like any ordinary hdd because it consists of one. The hdd cannot magically work faster so any data that is on it will be limited to the speed of the hdd. A Fusion Drive only has a very minor added bonus because it can move the data blocks that are used often from hdd to ssd so you have the full speed of an ssd. And that's also the disadvantage. If you want to have the speed of an ssd the data first needs to be moved from hdd to ssd. Keep in mind that the ssd is limited in size, just 128GB of which not all can be used for data (there is overhead for things like the partition, volume stuff, filesystem).
That's why a lot of things on the Fusion Drive are not faster than on an ordinary hdd. The added benefit of Fusion Drive over an ordinary hdd is just very limited. Most users won't tell the difference in machines like an iMac 27" because Apple uses 3.5" disks which are quite fast. If you want to safe money then stick to ordinary disks, buy the ssd if you want raw speed and the Fusion Drive if there isn't an option for an hdd. Fusion Drive works nice in benchmarks and for booting OS X (OS X will always be on the ssd which adds to the overhead, iirc the apps will also be on the ssd) but for real life use it doesn't add anything. If speed really doesn't matter all that much then getting the Fusion Drive is just as much a waste of money as the ssd option will be. The link to Anandtech posted by andy9l is a good overview of performance of a Fusion Drive. A previous post from roadkill401 has the same explanation as the Anandtech article.
Coming back to the Fusion Drive speed: since it moves blocks of data the speed is not the same. If you need something where the speed is the same or at least is somewhat predictable you can't use the Fusion Drive. An ordinary hdd is a better option and the ssd is the best in this case. When you do rendering, compiling and so on you want the speed to be predictable/continuous.
I'm sorry, but I completely disagree with what you just said. Small SSDs may or may not be better than Fusion Drives, but if they are, no one is giving good arguments for it here.
So are the Fusion Drive arguments. They are very weak and not thought out at all. No one seems to understand what a Fusion Drive actually is, why this causes problems and why because of that the ssd is more reliable. Only using sequential speeds in case of an ssd where it is iops, latency and random read/write speeds that actually matter when it comes to a desktop/laptop is another example. It's things like these that are missing.
Andy, you don't know what you are talking about.
Clearly neither do you...
A hard disk drive is simple. You tell it to write data somewhere, it writes it somewhere. If there is a crash, that bit of data may be destroyed, but 99.99% of your hard drive are just fine.
An SSD drive doesn't work like that. An SSD drive has to continuously find space where to write the data. And it has to remember what is written where. And that information is crucial for using the SSD drive, and when it is messed up, your SSD drive is history.
...because this is incorrect. Both an ssd and an hdd have to find where to put data and where data stored. Both of them store data in chunks on the drive. The exact way how they are doing it differs because it's different technology (magnetic vs electronic). The main reason why you have to defragment an hdd is because it stores data *somewhere*. The main reason why something like "seek time" is mentioned in hdd specs is because it stores data *somewhere*. And so on.
That's why enterprise drives have a huge capacitor so they can guarantee that they finish everything they started even if the power goes down.
No that's not why they do it. They do it to prevent dataloss or data corruption because the drive didn't write the data from its buffer (which is just RAM) to disk. If the power to the drive fails the buffer will be flushed causing dataloss. If you think about it, it's quite a gimmicky feature. It's nice that the buffer gets written to disk but that doesn't mean the data will be intact. If only part of the data was in the buffer only part of the data will be written to disk. Having only part of the data means data corruption/dataloss. Don't forget that in case of power loss this can also affect other components. For certain power loss scenarios just having an UPS is enough. The reason enterprise drives have things like that is to make them more reliable. It's about the entire feature set, not just 1 feature. And it isn't limited to an ssd, look up the main reason why RAID controllers have a BBU and what the main issue of RAID5 is.
Datarecovery for a Fusion Drive is also complicated since it is spanned across an ssd and an hdd AND data is moved between them in blocks. That means that one block can be on the hdd and the other on the ssd. Now think about what this means for datarecovery

Again, backups are quick and easy, datarecovery is only for when everything else failed.