The E4 has a usable working distance of about 200-300mm, but the TI10 and get down to about 20mm. This is great for debugging SMD designs, you can see 0402s and SOT323s glowing with ease, something that's not possible with the E4 and standard lens. As edavid says, the lens modifications involve adding a separate lens in front of the E4, and some people have reported success although I haven't tried it. You need to bugger about with 3D printing and ZnSe lenses from ebay, but there's LOTS of information in the other thread all about it. It's also possible to refocus the E4's lens by screwing it in (or out?) a bit to get closeup thermal shots. The E4 has a wider angle lens than the TI10 which explains the difference.
The quality of the Flir's thermal image is exceptional, much better than the TI10 in my opinion. But I got the TI10 in 2010 and paid about £3000 (~$5000) for it - the technology has moved on significantly since then.
My advice would be to demo each unit and see how you get on. Fluke were more than happy to send me a TI10 to try for a week for free, and I fell in love with it. I didn't demo the Flir, it was cheap enough that it wasn't necessary and I just jumped in. But even though the hacked E4 has higher resolution, is smaller and lighter and the battery lasts longer, I almost never use it in favour of the TI10 for the sort of electronics work we do.
Forum user 'Aurora' in the other thread seems to be somewhat of an expert on thermal imagers, he has a collection of them.