TLDR - the new system is far too complex and the requirements include support for cards not even used in the US.
In order to implement EMV aka chip & pin you need a device that is certified by EMVCO, and industry consortium. They issue LOAs (letters of authorization) for devices having passed the certification process. This administrative process is slow and expensive. Many device manufacturers have trouble getting their devices certified. Many of the devices you see in the marketplace may have chip reading hardware, but their firmware may not be up to date or certified. Certification is extremely complex due to the many variations of card and contactless support theoretically possible. There are two levels of certification needed. In short, the device manufacturers were not ready and the industry underfunded the certification authority. This is why proliferation of devices has been slow.
One you have a device whose firmware is certified the processing gateway and point of sale software has to be certified. This is an incredibly time consuming, expensive and arduous process.
There is a shortcut in this area known as a semi integrated solution. A pos implementer uses an already certified payment "black box" application to integrate with their POS system. This has many advantages but a big disadvantage. The semi integrated software is a middleman and in most cases exacts a price for the processing service making implementations of this approach less competitive.
Ideally systems will use a direct integration. This requires certification for all card brands and all card types. You need and expensive device called a Collis test tool to emulate every conceivable card and contactless technology type. There are hundreds of test cases for each card brand for all the possible scenarios, include failure fallback.
The problem is, the majority of these test cases are for cards never seen in the real world.
Chase issued chip and signature cards several years ago and the rest of the card brands realized that if they issued chip and pin cards, older folks and those who don't want to get pin numbers would use their Chase cards so all the card issuers went with chip and signature. Chip cards are hard to counterfeit (you have to be able to make the chips and I don't have a semiconductor foundry in my basement), but eliminates an important aspect of two factor authentication - something you know. Frankly chip and pin is better, but chip and signature is much better than what we have and probably good enough.
It will be another year before the backlog of certifications gets worked through. There is a waiting line to get slotted for certification and much of the time, the developers in line don't have what it takes to actually code the solution when its finally their turn. You don't google for solutions to these kind of problems. You really need to know exactly what you are doing. A developer of this kind of software cannot get it wrong and the software has to be defect free. And its very complex. If you are not experienced and you do not have a very high IQ and you are not willing to work extremely hard you don't have what it takes to write this king of code. This process is truly a bitch. Because the job is so big, the processing companies have offshored the certification liasons. Working through issues with offshore help protected by a bureaucracy is a special circle of hell reserved for those of us developers who must have done something heinous to deserve this fate...
As for the slowness of the new technology, there are a few factors that come into play. In the good old swipe world, the card is swiped and while the consumer is putting their card away, the device is getting an authorization in parallel. In the chip world, the consumer leaves the card in while the transaction is being processed. When the process is complete, they are asked to remove the card. This has the advantage of preventing consumers from forgetting their cards in the machine but has a big perceived performance hit. There is a technology called quick chip that allows the consumer to dip an remove their card and the processing occurs while they are putting the card away. This has a perceived performance advantage and we will all see most devices adopting this workflow over the next couple of years so the speed is improving. I for one am happy to wait a few extra seconds knowing that my card number cannot be stolen. The hassles of being on the phone with my card company explaining that I didn't make those charges far outweighs the few extra seconds.
Transitions like this are never easy but its worth it. The card brands (VISA, MasterCard, Amex, Discover, etc.) recognize that the certification process is too difficult and is simplifying the certification process.
To say the transition has been a disaster is an interesting statement. Its been a disaster for a relatively small few. Consumers have been impacted very little. Implementers have taken a huge hit and there will be the usual shakeout in the industry with small players that cannot get the job done. The card brands and issuers win big with lower fraud costs. The merchants are left holding the bag with increased equipment cost and greater exposure to fraud and disputes.